First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
概要
概要 (2)
研究の背景音源分離問題
研究の背景音源分離問題の種類
研究の背景歴史的発展
概要 (3)
ブラインド音源分離と独立成分分析
独立成分分析(ICA)おさらい
独立成分分析(ICA)おさらい (2)
独立成分分析(ICA)おさらい (3)
音源の分布音声
音源の分布ピアノ音
音源の分布ドラム音
耐残響性の向上周波数領域ICA(FDICA)
耐残響性の向上周波数領域ICA(FDICA) (2)
耐残響性の向上周波数領域ICA(FDICA) (3)
耐残響性の向上周波数領域ICA(FDICA) (4)
到来方向を用いたパーミュテーション解決
FDICAによる音源分離のメカニズム
FDICAの分離フィルタとABFの分離フィルタ
独立ベクトル分析(IVA)
IVAにおける音源分布と高次相関
IVAにおける音源分布と高次相関 (2)
FDICAとIVAの分離原理比較
概要 (4)
非負値行列因子分解の音響信号への適用
NMFのパラメータ推定
NMFの音響信号への応用例
NMFと音楽信号の相性
音楽信号は本当に低ランク
音楽信号は本当に低ランク (2)
板倉斎藤擬距離基準NMF(ISNMF)
板倉斎藤擬距離基準NMF(ISNMF) (2)
板倉斎藤擬距離基準NMF(ISNMF) (3)
板倉斎藤擬距離基準NMF(ISNMF) (4)
NMFの多チャネル信号への拡張
NMFの多チャネル信号への拡張 (2)
空間相関行列とは
空間相関行列のランク
NMFの多チャネル信号の最適化
概要 (5)
動機
様々な非ガウス分布を仮定したIVA
「低ランク性」の音源モデルへの導入
提案手法IVAとNMFを融合した新しいBSS
提案手法IVAとNMFを融合した新しいBSS (2)
ILRMAのコスト関数と潜在変数の導入
ILRMAの最適化
ILRMAの更新のイメージ
多チャネルNMFとILRMAの関連性
多チャネルNMFとILRMAの関連性 (2)
多チャネルNMFとILRMAの関連性 (3)
IVA多チャネルNMFILRMAの関連性
尤度関数の比較
歴史的発展(再掲)
従来手法とILRMAの性能評価
音楽音源分離実験の条件
実験結果 fort_minor-remember_the_name
実験結果 ultimate_nz_tour
各手法の性能と計算コストの比較例
音声音源分離実験の条件
実験結果 female3_liverec_1m
実験結果 male3_liverec_1m
各手法の性能の主観評価による比較
考察
音源分離デモンストレーション音楽分離の例
概要 (6)
より高精度なブラインド音源分離を目指して
参考文献(アルファベット順)(14)
参考文献(アルファベット順)(24)
参考文献(アルファベット順)(34)
参考文献(アルファベット順)(44)
19
到来方向を用いたパーミュテーション解決bull FDICA+ DOA クラスタリング [Saruwatari 2006]
ndash 推定分離フィルタ から混合フィルタ を逆算ndash 音源の到来方向( DOA )でクラスタリング
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
First speaker in female speech signal (dev1) Second speaker in female speech signal (dev1) Guitar in music signal (song ID4) Synth in music signal (song ID4)
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
ndash では空間情報の学習データは容易可能かbull 録音環境は一期一会bull 学習なし or スモールデータの追求
時間区間の指定
領域の指定
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
概要
概要 (2)
研究の背景音源分離問題
研究の背景音源分離問題の種類
研究の背景歴史的発展
概要 (3)
ブラインド音源分離と独立成分分析
独立成分分析(ICA)おさらい
独立成分分析(ICA)おさらい (2)
独立成分分析(ICA)おさらい (3)
音源の分布音声
音源の分布ピアノ音
音源の分布ドラム音
耐残響性の向上周波数領域ICA(FDICA)
耐残響性の向上周波数領域ICA(FDICA) (2)
耐残響性の向上周波数領域ICA(FDICA) (3)
耐残響性の向上周波数領域ICA(FDICA) (4)
到来方向を用いたパーミュテーション解決
FDICAによる音源分離のメカニズム
FDICAの分離フィルタとABFの分離フィルタ
独立ベクトル分析(IVA)
IVAにおける音源分布と高次相関
IVAにおける音源分布と高次相関 (2)
FDICAとIVAの分離原理比較
概要 (4)
非負値行列因子分解の音響信号への適用
NMFのパラメータ推定
NMFの音響信号への応用例
NMFと音楽信号の相性
音楽信号は本当に低ランク
音楽信号は本当に低ランク (2)
板倉斎藤擬距離基準NMF(ISNMF)
板倉斎藤擬距離基準NMF(ISNMF) (2)
板倉斎藤擬距離基準NMF(ISNMF) (3)
板倉斎藤擬距離基準NMF(ISNMF) (4)
NMFの多チャネル信号への拡張
NMFの多チャネル信号への拡張 (2)
空間相関行列とは
空間相関行列のランク
NMFの多チャネル信号の最適化
概要 (5)
動機
様々な非ガウス分布を仮定したIVA
「低ランク性」の音源モデルへの導入
提案手法IVAとNMFを融合した新しいBSS
提案手法IVAとNMFを融合した新しいBSS (2)
ILRMAのコスト関数と潜在変数の導入
ILRMAの最適化
ILRMAの更新のイメージ
多チャネルNMFとILRMAの関連性
多チャネルNMFとILRMAの関連性 (2)
多チャネルNMFとILRMAの関連性 (3)
IVA多チャネルNMFILRMAの関連性
尤度関数の比較
歴史的発展(再掲)
従来手法とILRMAの性能評価
音楽音源分離実験の条件
実験結果 fort_minor-remember_the_name
実験結果 ultimate_nz_tour
各手法の性能と計算コストの比較例
音声音源分離実験の条件
実験結果 female3_liverec_1m
実験結果 male3_liverec_1m
各手法の性能の主観評価による比較
考察
音源分離デモンストレーション音楽分離の例
概要 (6)
より高精度なブラインド音源分離を目指して
参考文献(アルファベット順)(14)
参考文献(アルファベット順)(24)
参考文献(アルファベット順)(34)
参考文献(アルファベット順)(44)
70
参考文献(アルファベット順)( 14 )bull [Araki1 2003] S Araki R Mukai S Makino T Nishikawa and H Saruwatari ldquoThe
fundamental limitation of frequency domain blind source separation for convolutive mixtures of speechrdquo IEEE Trans Speech and Audio Process vol 11 no 2 pp 109ndash116 2003
bull [Araki2 2003] S Araki S Makino Y Hinamoto R Mukai T Nishikawa and H Saruwatari ldquoEquivalence between frequency-domain blind source separation and frequency-domain adaptive beamforming for convolutive mixturesrdquo EURASIP Journal on Advances in Signal Process vol 2003 no 11 pp 1ndash10 2003
bull [Comon 1994] P Comon ldquoIndependent component analysis a new conceptrdquo Signal Process vol 36 no 3 pp 287ndash314 1994
bull [Duong 2010] N Q K Duong E Vincent and R Gribonval ldquoUnder-determined reverberant audio source separation using a full-rank spatial covariance modelrdquo IEEE Trans Audio Speech Lang Process vol 18 no 7 pp 1830ndash1840 2010
bull [Feacutevotte 2009] C Feacutevotte N Bertin and J-LDurrieu ldquoNonnegative matrix factorization with the Itakura-Saito divergence With application to music analysisrdquo Neural Comput vol 21 no 3 pp 793ndash830 2009
bull [Hiroe 2006] A Hiroe ldquoSolution of permutation problem in frequency domain ICA using multivariate probability density functionsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 601ndash608
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
概要
概要 (2)
研究の背景音源分離問題
研究の背景音源分離問題の種類
研究の背景歴史的発展
概要 (3)
ブラインド音源分離と独立成分分析
独立成分分析(ICA)おさらい
独立成分分析(ICA)おさらい (2)
独立成分分析(ICA)おさらい (3)
音源の分布音声
音源の分布ピアノ音
音源の分布ドラム音
耐残響性の向上周波数領域ICA(FDICA)
耐残響性の向上周波数領域ICA(FDICA) (2)
耐残響性の向上周波数領域ICA(FDICA) (3)
耐残響性の向上周波数領域ICA(FDICA) (4)
到来方向を用いたパーミュテーション解決
FDICAによる音源分離のメカニズム
FDICAの分離フィルタとABFの分離フィルタ
独立ベクトル分析(IVA)
IVAにおける音源分布と高次相関
IVAにおける音源分布と高次相関 (2)
FDICAとIVAの分離原理比較
概要 (4)
非負値行列因子分解の音響信号への適用
NMFのパラメータ推定
NMFの音響信号への応用例
NMFと音楽信号の相性
音楽信号は本当に低ランク
音楽信号は本当に低ランク (2)
板倉斎藤擬距離基準NMF(ISNMF)
板倉斎藤擬距離基準NMF(ISNMF) (2)
板倉斎藤擬距離基準NMF(ISNMF) (3)
板倉斎藤擬距離基準NMF(ISNMF) (4)
NMFの多チャネル信号への拡張
NMFの多チャネル信号への拡張 (2)
空間相関行列とは
空間相関行列のランク
NMFの多チャネル信号の最適化
概要 (5)
動機
様々な非ガウス分布を仮定したIVA
「低ランク性」の音源モデルへの導入
提案手法IVAとNMFを融合した新しいBSS
提案手法IVAとNMFを融合した新しいBSS (2)
ILRMAのコスト関数と潜在変数の導入
ILRMAの最適化
ILRMAの更新のイメージ
多チャネルNMFとILRMAの関連性
多チャネルNMFとILRMAの関連性 (2)
多チャネルNMFとILRMAの関連性 (3)
IVA多チャネルNMFILRMAの関連性
尤度関数の比較
歴史的発展(再掲)
従来手法とILRMAの性能評価
音楽音源分離実験の条件
実験結果 fort_minor-remember_the_name
実験結果 ultimate_nz_tour
各手法の性能と計算コストの比較例
音声音源分離実験の条件
実験結果 female3_liverec_1m
実験結果 male3_liverec_1m
各手法の性能の主観評価による比較
考察
音源分離デモンストレーション音楽分離の例
概要 (6)
より高精度なブラインド音源分離を目指して
参考文献(アルファベット順)(14)
参考文献(アルファベット順)(24)
参考文献(アルファベット順)(34)
参考文献(アルファベット順)(44)
71
参考文献(アルファベット順)( 24 )bull [James 1961] W James and C Stein ldquoEstimation with quadratic lossrdquo Proc Berkeley
Symposium on Mathematical Statistics and Probability vol 1 1961 pp 361ndash379bull [Kim 2006] T Kim T Eltoft and T-W Lee ldquoIndependent vector analysis An extension of
ICA to multivariate componentsrdquo Proc Int Conf Independent Compon Anal Blind Source Separation 2006 pp 165ndash172
bull [Kim 2007] T Kim H T Attias S-Y Lee and T-W Lee ldquoBlind source separation exploiting higher-order frequency dependenciesrdquo IEEE Trans Audio Speech Lang Process vol 15 no 1 pp 70ndash79 2007
bull [Kitamura1 2014] D Kitamura H Saruwatari K Yagi K Shikano Y Takahashi and K Kondo ldquoMusic signal separation based on supervised nonnegative matrix factorization with orthogonality and maximum-divergence penaltiesrdquo IEICE Trans Fundamentals of Electronics Communications and Computer Sciences vol E97-A no 5 pp 1113ndash1118 2014
bull [Kitamura2 2014] T Miyauchi D Kitamura H Saruwatari and S Nakamura ldquoDepth estimation of sound images using directional clustering and activation-shared nonnegative matrix factorizationrdquo Journal of Signal Process vol 18 no 4 pp 217ndash220 2014
bull [Kitamura 2015] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S Nakamura ldquoMultichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restorationrdquo IEEEACM Trans on Audio Speech and Lang Process vol 23 no 4 pp 654ndash669 2015
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
概要
概要 (2)
研究の背景音源分離問題
研究の背景音源分離問題の種類
研究の背景歴史的発展
概要 (3)
ブラインド音源分離と独立成分分析
独立成分分析(ICA)おさらい
独立成分分析(ICA)おさらい (2)
独立成分分析(ICA)おさらい (3)
音源の分布音声
音源の分布ピアノ音
音源の分布ドラム音
耐残響性の向上周波数領域ICA(FDICA)
耐残響性の向上周波数領域ICA(FDICA) (2)
耐残響性の向上周波数領域ICA(FDICA) (3)
耐残響性の向上周波数領域ICA(FDICA) (4)
到来方向を用いたパーミュテーション解決
FDICAによる音源分離のメカニズム
FDICAの分離フィルタとABFの分離フィルタ
独立ベクトル分析(IVA)
IVAにおける音源分布と高次相関
IVAにおける音源分布と高次相関 (2)
FDICAとIVAの分離原理比較
概要 (4)
非負値行列因子分解の音響信号への適用
NMFのパラメータ推定
NMFの音響信号への応用例
NMFと音楽信号の相性
音楽信号は本当に低ランク
音楽信号は本当に低ランク (2)
板倉斎藤擬距離基準NMF(ISNMF)
板倉斎藤擬距離基準NMF(ISNMF) (2)
板倉斎藤擬距離基準NMF(ISNMF) (3)
板倉斎藤擬距離基準NMF(ISNMF) (4)
NMFの多チャネル信号への拡張
NMFの多チャネル信号への拡張 (2)
空間相関行列とは
空間相関行列のランク
NMFの多チャネル信号の最適化
概要 (5)
動機
様々な非ガウス分布を仮定したIVA
「低ランク性」の音源モデルへの導入
提案手法IVAとNMFを融合した新しいBSS
提案手法IVAとNMFを融合した新しいBSS (2)
ILRMAのコスト関数と潜在変数の導入
ILRMAの最適化
ILRMAの更新のイメージ
多チャネルNMFとILRMAの関連性
多チャネルNMFとILRMAの関連性 (2)
多チャネルNMFとILRMAの関連性 (3)
IVA多チャネルNMFILRMAの関連性
尤度関数の比較
歴史的発展(再掲)
従来手法とILRMAの性能評価
音楽音源分離実験の条件
実験結果 fort_minor-remember_the_name
実験結果 ultimate_nz_tour
各手法の性能と計算コストの比較例
音声音源分離実験の条件
実験結果 female3_liverec_1m
実験結果 male3_liverec_1m
各手法の性能の主観評価による比較
考察
音源分離デモンストレーション音楽分離の例
概要 (6)
より高精度なブラインド音源分離を目指して
参考文献(アルファベット順)(14)
参考文献(アルファベット順)(24)
参考文献(アルファベット順)(34)
参考文献(アルファベット順)(44)
72
参考文献(アルファベット順)( 34 )bull [Kitamura 2016] D Kitamura H Saruwatari H Kameoka Y Takahashi K Kondo and S
Nakamura ldquoDetermined blind source separation unifying independent vector analysis and nonnegative matrix factorizationrdquo IEEEACM Trans Audio Speech Lang Process vol 24 no 9 pp 1626ndash1641 2016
bull [Kulis 2006] B Kulis M Sustik and I Dhillon ldquoLearning low-rank kernel matricesrdquo Proc Int Conf on Machine Learning 2006 pp 505ndash512
bull [Lee 1999] D D Lee and H S Seung ldquoLearning the parts of objects by non-negative matrix factorizationrdquo Nature vol 401 pp 788ndash791 1999
bull [Lee 2000] D D Lee and H S Seung ldquoAlgorithms for non-negative matrix factorizationrdquo Proc Adv Neural Inform Process Syst 2000 vol 13 pp 556ndash562
bull [Ono 2011] N Ono ldquoStable and fast update rules for independent vector analysis based on auxiliary function techniquerdquo Proc IEEE Workshop on Applications of Signal Process to Audio and Acoust 2011 pp 189ndash192
bull [Ono 2012] T Ono N Ono and S Sagayama ldquoUser-guided independent vector analysis with source activity tuningrdquo Proc ICASSP 2012 pp 2417ndash2420
bull [Ozerov 2010] A Ozerov and C Feacutevotte ldquoMultichannel nonnegative matrix factorization in convolutive mixtures for audio source separationrdquo IEEE Trans Audio Speech and Lang Process vol 18 no 3 pp 550ndash563 2010
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b
概要
概要 (2)
研究の背景音源分離問題
研究の背景音源分離問題の種類
研究の背景歴史的発展
概要 (3)
ブラインド音源分離と独立成分分析
独立成分分析(ICA)おさらい
独立成分分析(ICA)おさらい (2)
独立成分分析(ICA)おさらい (3)
音源の分布音声
音源の分布ピアノ音
音源の分布ドラム音
耐残響性の向上周波数領域ICA(FDICA)
耐残響性の向上周波数領域ICA(FDICA) (2)
耐残響性の向上周波数領域ICA(FDICA) (3)
耐残響性の向上周波数領域ICA(FDICA) (4)
到来方向を用いたパーミュテーション解決
FDICAによる音源分離のメカニズム
FDICAの分離フィルタとABFの分離フィルタ
独立ベクトル分析(IVA)
IVAにおける音源分布と高次相関
IVAにおける音源分布と高次相関 (2)
FDICAとIVAの分離原理比較
概要 (4)
非負値行列因子分解の音響信号への適用
NMFのパラメータ推定
NMFの音響信号への応用例
NMFと音楽信号の相性
音楽信号は本当に低ランク
音楽信号は本当に低ランク (2)
板倉斎藤擬距離基準NMF(ISNMF)
板倉斎藤擬距離基準NMF(ISNMF) (2)
板倉斎藤擬距離基準NMF(ISNMF) (3)
板倉斎藤擬距離基準NMF(ISNMF) (4)
NMFの多チャネル信号への拡張
NMFの多チャネル信号への拡張 (2)
空間相関行列とは
空間相関行列のランク
NMFの多チャネル信号の最適化
概要 (5)
動機
様々な非ガウス分布を仮定したIVA
「低ランク性」の音源モデルへの導入
提案手法IVAとNMFを融合した新しいBSS
提案手法IVAとNMFを融合した新しいBSS (2)
ILRMAのコスト関数と潜在変数の導入
ILRMAの最適化
ILRMAの更新のイメージ
多チャネルNMFとILRMAの関連性
多チャネルNMFとILRMAの関連性 (2)
多チャネルNMFとILRMAの関連性 (3)
IVA多チャネルNMFILRMAの関連性
尤度関数の比較
歴史的発展(再掲)
従来手法とILRMAの性能評価
音楽音源分離実験の条件
実験結果 fort_minor-remember_the_name
実験結果 ultimate_nz_tour
各手法の性能と計算コストの比較例
音声音源分離実験の条件
実験結果 female3_liverec_1m
実験結果 male3_liverec_1m
各手法の性能の主観評価による比較
考察
音源分離デモンストレーション音楽分離の例
概要 (6)
より高精度なブラインド音源分離を目指して
参考文献(アルファベット順)(14)
参考文献(アルファベット順)(24)
参考文献(アルファベット順)(34)
参考文献(アルファベット順)(44)
73
参考文献(アルファベット順)( 44 )bull [Saruwatari 2000] S Kurita H Saruwatari S Kajita K Takeda and F Itakura ldquoEvaluation
of blind signal separation method using directivity pattern under reverberant conditionsrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2000 pp 3140ndash3143
bull [Saruwatari 2006] H Saruwatari T Kawamura T Nishikawa A Lee and K Shikano ldquoBlind source separation based on a fast-convergence algorithm combining ICA and beamformingrdquo IEEE Trans Audio Speech Lang Process vol 14 no 2 pp 666ndash678 Mar 2006
bull [Sawada 2004] H Sawada R Mukai S Araki and SMakino ldquoConvolutive blind source separation for more than two sources in the frequency domainrdquo Proc IEEE Int Conf Acoust Speech Signal Process 2004 pp III-885ndashIII-888
bull [Sawada 2013] H Sawada HKameoka SAraki and N Ueda ldquoMultichannel extensions of non-negative matrix factorization with complex-valued datardquo IEEE Trans Audio Speech Lang Process vol 21 no 5 pp 971ndash982 2013
bull [Smaragdis 1998] P Smaragdis ldquoBlind separation of convolved mixtures in the frequency domainrdquo Neurocomputing vol 22 pp 21ndash34 1998
bull [Smaragdis 2007] P Smaragdis B Raj and M Shashanka ldquoSupervised and semi-supervised separation of sounds from single-channel mixturesrdquo Proc ICA 2007 pp 414ndash421
独立性に基づくブラインド音源分離の発展と 独立低ランク行列分析 History of independence-based b