lecture notes in computer science 10597 - springer978-3-319-69900-4/1.pdf · lecture notes in...
TRANSCRIPT
Lecture Notes in Computer Science 10597
Commenced Publication in 1973Founding and Former Series Editors:Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David HutchisonLancaster University, Lancaster, UK
Takeo KanadeCarnegie Mellon University, Pittsburgh, PA, USA
Josef KittlerUniversity of Surrey, Guildford, UK
Jon M. KleinbergCornell University, Ithaca, NY, USA
Friedemann MatternETH Zurich, Zurich, Switzerland
John C. MitchellStanford University, Stanford, CA, USA
Moni NaorWeizmann Institute of Science, Rehovot, Israel
C. Pandu RanganIndian Institute of Technology, Madras, India
Bernhard SteffenTU Dortmund University, Dortmund, Germany
Demetri TerzopoulosUniversity of California, Los Angeles, CA, USA
Doug TygarUniversity of California, Berkeley, CA, USA
Gerhard WeikumMax Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7412
B. Uma Shankar • Kuntal GhoshDeba Prasad Mandal • Shubhra Sankar RayDavid Zhang • Sankar K. Pal (Eds.)
Pattern Recognitionand Machine Intelligence7th International Conference, PReMI 2017Kolkata, India, December 5–8, 2017Proceedings
123
EditorsB. Uma ShankarIndian Statistical InstituteKolkataIndia
Kuntal GhoshIndian Statistical InstituteKolkataIndia
Deba Prasad MandalIndian Statistical InstituteKolkataIndia
Shubhra Sankar RayIndian Statistical InstituteKolkataIndia
David ZhangThe Hong Kong Polytechnic UniversityHong KongChina
Sankar K. PalIndian Statistical InstituteKolkataIndia
ISSN 0302-9743 ISSN 1611-3349 (electronic)Lecture Notes in Computer ScienceISBN 978-3-319-69899-1 ISBN 978-3-319-69900-4 (eBook)https://doi.org/10.1007/978-3-319-69900-4
Library of Congress Control Number: 2017957548
LNCS Sublibrary: SL6 – Image Processing, Computer Vision, Pattern Recognition, and Graphics
© Springer International Publishing AG 2017This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of thematerial is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting, reproduction on microfilms or in any other physical way, and transmission or informationstorage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology nowknown or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoes not imply, even in the absence of a specific statement, that such names are exempt from the relevantprotective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in this book arebelieved to be true and accurate at the date of publication. Neither the publisher nor the authors or the editorsgive a warranty, express or implied, with respect to the material contained herein or for any errors oromissions that may have been made. The publisher remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer NatureThe registered company is Springer International Publishing AGThe registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
It is a great pleasure to introduce you all to the proceedings of the 7th InternationalConference on Pattern Recognition and Machine Intelligence (PReMI 2017), held atthe Indian Statistical Institute (ISI), Kolkata, India, during December 5–8, 2017. Theobjective of the conference is to introduce to the community the most recentadvancements in research in the domain of pattern recognition and machine intelli-gence. Our goal is to encourage academic and industrial collaboration in all relatedfields in machine learning involving scientists, engineers, professionals, researchers,and students from India and abroad. The conference is held biennially to make it anideal platform for researcher all over the world to come and share their views andexperiences. This was the seventh edition in this series, being held in the year markingthe 125th birthday of late Prof. Prasanta Chandra Mahalanobis.
Professor Mahalanobis was the founder of the Indian Statistical Institute and thefather of modern statistics in India. As researchers in pattern recognition and machinelearning we are immensely indebted to him. He was instrumental in inspiring thedesign of the first analog computer in India in 1953. He brought to ISI the first digitalcomputer to India in the year 1955. As a mark of our respect to this monumentalpersonality, we organized a Special Session on “Celebration of 125th Birth Anniver-sary of Professor P.C. Mahalanobis” at PReMI 2017.
The conference comprised several keynote and invited lecturers delivered by emi-nent and distinguished researchers from around the world. Both the invited and thetechnical sessions featured interesting lectures in classic and contemporary aspects ofmachina intelligence. The topics range from deep learning and Internet of Things(IoT) to computer vision and big data analytics. There were two exclusive sessions on“Deep Learning” and “Spatial Data Science and Engineering” Like previous editions,PReMI 2017 had a very good response in terms of paper submissions. Altogether therewere 293 submissions from about 15 countries spanning three continents. Each paperwas critically reviewed by experts in the field, after which 85 papers (29% acceptancerate) were accepted for inclusion in these proceedings. The accepted papers are dividedinto ten groups, although there could be some overlap. Articles written by the keynoteand invited speakers are also included in the proceedings (mostly abstracts).
We wish to express our appreciation to the Program Committee and TechnicalReview Committee members, who worked hard to ensure the quality of the contri-butions of this volume. We are thankful to the editors of the journals FundamentaInformaticae and Applied Soft Computing for kindly agreeing to publish the extendedversions of some of the selected papers in their esteemed journals. We also take thisopportunity to thank Professors Vineet Bafna, Andrzej Skowron, Farzin Deravi,Upinder S. Bhalla, Uday B. Desai, Soumen Chakraborti, Ambarish Ghosh, ParthaPratim Majumder, Probal Chaudhuri, Subhasis Chaudhuri, David Zhang, and ShalabhBhatnagar for accepting our invitation to deliver keynote, invited, and special lecturesduring the conference. We gratefully acknowledge Alfred Hofmann of Springer for his
co-operation in the publication of the PReMI 2017 proceedings in the LNCS series, asdone for the previous editions. We would like to thank all the organizations who eitherendorsed or sponsored this conference technically or financially. We are grateful toEasyChair for providing us with a wonderful platform for conducting the entire processof paper review. Last but not the least, we take this opportunity to thank all thecontributors for their enthusiastic response, without which no conference can ever besuccessful.
While preparing the proceedings we mourned the sad demise of Professor Lotfi A.Zadeh, the founder of fuzzy mathematics, an imperative part of contemporary machinelearning. He was on the advisory board of PReMI ever since its inception in 2005,including the present edition. Our institute honored him with a doctor honoris causa in2006 during its annual convocation. We express our deep condolences to his familyand all his friends and colleagues. It is a great loss to the pattern recognition and softcomputing/computational intelligence community.
Our best wishes to all the participants of PReMI 2017 conference. May this volume,which contains the papers presented at PReMI 2017 prove to be a valuable source ofreference for ongoing and future research work.
December 2017 B. Uma ShankarKuntal Ghosh
Deba Prasad MandalShubhra Sankar Ray
David ZhangSankar K. Pal
VI Preface
Message from the General Chair
PReMI, the biennial International Conference on Pattern Recognition and MachineIntelligence, returned to Kolkata, the City of Joy, after its sixth edition in Warsaw,Poland, in June/July 2015! I am delighted that the seventh edition (PReMI 2017) washeld in the year that marks the 125th birthday of late Prof. Prasanta Chandra Maha-lanobis, the founder of our Indian Statistical Institute.
Like earlier versions, PReMI 2017 had a nice mixture of keynote and invitedspeeches, and quality research papers using both classic and modern computingparadigms, covering different facets of pattern recognition and machine intelligencewith real-life applications. Apart from classic topics, special emphasis was given tocontemporary research areas such as big data analytics, deep learning, Internet ofThings, and computer vision through both regular and special sessions. Somepost-conference special issues will be published as done in the past. All these makePReMI 2017 an ideal state-of-the-art platform for researchers and practitioners toexchange ideas and enrich their knowledge.
I thank all the participants, speakers, reviewers, and members of various committeesfor making this event a grand success. My thanks are also due to the sponsors for theirsupport, and Springer for publishing the PReMI proceedings, since its first edition in2005, in the prestigious LNCS series.
I trust, the participants had an academically fruitful and enjoyable stay in Kolkata.
December 2017 Sankar K. Pal
Organization
PReMI 2017 was organized by the Machine Intelligence Unit, Indian StatisticalInstitute (ISI) in Kolkata during December 5–8, 2017.
Conference Committee
Patron
SanghamitraBandyopadhyay
ISI, Kolkata, India
General Chair
Sankar K. Pal ISI, Kolkata, India
Program Chairs
David Zhang PolyU, Hong Kong, SAR ChinaKuntal Ghosh ISI, Kolkata, IndiaB. Uma Shankar ISI, Kolkata, India
Organizing Chairs
Deba Prasad Mandal ISI, Kolkata, IndiaShubhra Sankar Ray ISI, Kolkata, India
Special Session Chairs
Ashish Ghosh ISI, Kolkata, IndiaFarid Melgani University of Trento, ItalySambhunath Biswas TIU, Kolkata, IndiaAlfredo Petrosino University of Naples, ItalySoumya K. Ghosh IIT Kharagpur, India
Special Issue Chairs
Sushmita Mitra ISI, Kolkata, IndiaP.N. Suganthan NTU, Singapore
Industry Liaisons
Malay K. Kundu ISI, Kolkata, IndiaC.A. Murthy ISI, Kolkata, IndiaSantanu Chaudhury CEERI, Pilani, India
International Liaisons
Sergei O. Kuznetsov HSE, Moscow, RussiaMarzena Kryszkiewicz WUT, PolandSimon C.K. Shiu PolyU, Hong Kong, SAR China
Advisory Committee
Lofti A. Zadeh, USAC.R. Rao, USAAnil K. Jain, USAJosef Kittler, UKLaveen N. Kanal, USAB.L. Deekshatulu, IndiaDwijesh Dutta Majumder, IndiaAndrzej Skowron, PolandRama Chellappa, USAWitold Pedrycz, CanadaDavid W. Aha, USAGabriella Sanniti di Baja, ItalyB. Yegnanarayana, IndiaShun-ichi Amari, JapanJayaram Udupa, USAJiming Liu, Hong Kong, SAR ChinaRonald Yager, USANing Zhong, JapanTharram Dillon, AustraliaHenryk Rybinski, Poland
Program Committee
Jayadeva Indian Institute of Technology Delhi, IndiaTinku Acharya Videonetics Technology Pvt. Ltd., IndiaMd. Atiqur Rahman Ahad University of Dhaka, BangladeshMohua Banerjee Indian Institute of Technology Kanpur, IndiaJayanta Basak NetApp, IndiaSmarajit Bose Indian Statistical Institute, IndiaRoberto M. Cesar USP, BrazilGoutam Chakraborty Iwate Prefectural University, JapanBhabatosh Chanda Indian Statistical Institute, IndiaSubhasis Chaudhuri Indian Institute of Technology Bombay, IndiaSung-Bae Cho Yonsei University, South KoreaPartha Pratim Das Indian Institute of Technology Kharagpur, IndiaSukhendu Das Indian Institute of Technology Madras, IndiaDipankar Dasgupta The University of Memphis, USARajat K. De Indian Statistical Institute, India
X Organization
Farzin Deravi University of Kent, UKShaikh A. Fattah BUET, BangladeshPaolo Gamba University of Pavia, ItalyJoydeep Ghosh University of Texas, USAMark Girolami University of Warwick, UKPhalguni Gupta NITTTR Kolkata, IndiaLarry Hall University of South Florida, USAFrancisco Herrera University of Granada, SpainQinghua Hu Tianjin University, ChinaC.V. Jawahar IIIT Hyderabad, IndiaJohn Kerekes Rochester Institute of Technology, USARavi Kothari IBM Research, IndiaPawan Lingras Saint Mary’s University, USAPradipta Maji Indian Statistical Institute, IndiaFrancesco Masulli University of Genoa, ItalyPabitra Mitra Indian Institute of Technology Kharagpur, IndiaJayanta Mukherjee Indian Institute of Technology Kharagpur, IndiaM.N. Murty Indian Institute of Science, IndiaY. Narahari Indian Institute of Science, IndiaB.L. Narayan Yahoo! Labs, USANasser Nasrabadi West Virginia University, USANikhil Rajan Pal Indian Statistical Institute, IndiaAngel P. Del Pobil Universitat Jaume I, SpainAmit K. Roy-Chowdhury University of California, USAPunam Saha University of Iowa, USAP.S. Sastry Indian Institute of Science, IndiaFaisal Shafait National University of Sciences and Technology,
PakistanSitabhra Sinha The Institute of Mathematical Sciences, IndiaDominik Slezak University of Warsaw, PolandBrijesh Verma Central Queensland University, AustraliaDianhui Wang La Trobe University, Australia
Technical Review Committee
Anand, AshishBagchi, AdityaBanerjee, AbhirupBanerjee, MinakshiBanerjee, RomiBanerjee, SubhashisBanerjee, SwatiBanka, HaiderBasu, TanmayBhadra, Tapas
Bhandari, DinabandhuBhattacharya, BhargabBhattacharya, UjjwalBhattacharyya, BalaramBhattacharyya, Dhruba K.Bhattacharyya, MalayBishnu, Partha SarathiChaki, NabenduChakrabarty, AbhisekChakraborty, Debarati
Chakraborty, DebasritaChatterjee, GargaChattopadhyay,TanushyamChellu, Chandra SekharChen, BoChowdhury, Ananda S.Chowdhury, ManishDas, ApurbaDas, Chandra
Organization XI
Das, KoelDas, MonidipaDas, SaurabhDas, SudebDasgupta, ArindamDatta, AlokeDehuri, SatchidanandaDey, BhaskarDutta, ParamarthaEkbal, AsifGarain, UtpalHalder, AnindyaKundu, SumanKuppili,
VenkatanareshbabuLaw, AnweshaMaitra, SanjitMajumdar, DebapriyoMandal, AnkitaMazumdar, DebasisMeher, Saroj K.Mishra, DeepakMisra, SudipMitra, Mandar
Mitra, SumanMitra, Suman KumarMittal, NamitaMohanta, Partha PratimMondal, AjoyMukherjee, Dipti PrasadMukhopadhyay, SubhasisMurthy, K. RamachandraNaik, SarifNanda, Pradipta KumarNaresh, K.M.Nayak, LosianaPal, Jayanta KumarPal, MonalisaPal, RajarshiPal, Rajat KumarPal, UmapadaParui, Swapan KumarPatil, HemantPatra, SwarnajyotiPaul, GoutamPaul, SushmitaPhadikar, AmitPrasad, M.V.N.K.
Prasanna, S.R.M.Ray, SumantaReddy, DamodarRoy, MonideepaRoy, RahulRoy, ShaswatiSa, Pankaj K.Saha, Sanjoy KumarSanyal, Debarshi KumarSen, DebashisSenapati, ApurbalalShah, EktaSharma, AnmolShrein, John M.Sil, JayaSinha, DebajyotiSubrahmanyam, GorthiSubudhi, Badri NarayanTrzcinski, TomaszVeerakumar, T.Verma, ManishaZaveri, MukeshZhang, Hao
Sponsoring Organizations
Endorsed by
International Association for Pattern Recognition (IAPR)
Technical Co-sponsor
IEEE Kolkata Section
Other Sponsors
Center for Soft Computing Research: A National Facility, ISI, KolkataWeb Intelligence Consortium (WIC)International Rough Set Society (IRSS)INAE Kolkata ChapterWorld Federation on Soft Computing (WFSC)Springer International Publishing
XII Organization
Interactive Granular Computingin Data Science
Andrzej Skowron1, 2
1 Faculty of Mathematics, Computer Science and Mechanics,University of Warsaw, [email protected]
2 Systems Research Institute, Polish Academy of Sciences
We discuss Interactive Granular Computing (IGrC) as the basis of a Data Sciencecomputing model. IGrC binds together and brings a synchronous cooperation amongthe following four basic concepts of Artificial Intelligence: language, reasoning, per-ception, and action. This, together with information granulation, helps agents to dealwith many complex tasks of perceiving or transforming compound abstract andphysical objects (e.g., in the context of complex spatio-temporal space). One shouldconsider that in Data Science agents collecting data have control over the dataacquisition, i.e., they are deciding say which data, using which sources, at what time,and why should be collected.
Basic objects in IGrC are complex granules (c-granules or granules, for short).They are grounded in the physical reality and are, in particular, responsible for gen-eration of the networks of information systems (data tables) through interactions withthe configurations of physical objects. Development of a particular network of infor-mation systems is guided by the need to learn the relevant computational buildingblocks that are necessary for perception, using the formulation by Leslie Valiant.Among these blocks, often learned hierarchically, one can distinguish patterns, clustersor classifiers. The computational building blocks are used by agents, e.g., forapproximation of conditions responsible for initiating actions or plans. Agents per-forming computations based on interaction with the physical environment learn newc-granules, in particular, in the form of interaction rules, representing knowledge notknown a priori by agents. These new c-granules are used not only for construction ofcompound abstract objects but also of compound physical objects, e.g., sensors com-posed out of more primitive sensors. Learning of interaction rules also supports thecontrol of agents, in particular the self-organized distributed control. Numerous tasks ofagents may be classified as control tasks performed by agents aiming at achieving thehigh quality computational trajectories of configurations of c-granules relative to theconsidered quality measures over the trajectories.
Reasoning supporting agents in searching for solutions of their tasks is based onadaptive judgment, an important component of IGrC. Methods based on adaptivejudgment allow agents to construct from given configurations of their c-granules newones. These new configurations of c-granules should be constructed taking into accountthe needs of agents realized through interactions with the environment. Here, new
challenges are related to developing strategies for predicting and controlling behaviorsof agents. We propose to investigate these challenges using the IGrC framework withadaptive judgment used for controlling of computations performed on c-granules. Forexample, adaptive judgment is used in adaptive learning of rough set based approxi-mations of complex vague concepts evolving with time. It is also used in the riskmanagement of granular computations, carried out by agents, toward achieving theagent needs.
XVI A. Skowron
Identifying the Favored Allelein a Selective Sweep
Vineet Bafna
Computer Science and Engineering University of California,San Diego, USA
Abstract. Selection is a dominant force in evolution. Mutations arising at ran-dom might favor individuals in a specific environmental niche, and populationsadapt by rapidly increasing the frequency of individuals carrying the favoredmutations. The selection process results in distinct patterns (a signature) of allelefrequencies and haplotype structures that can be exploited to identify the genesresponding to selection pressure. A study of selection signals in humans has ledto molecular insight into the evolution of many natural traits such as skin andeye color, as also adaptation to extreme environments.
Computational methods that scan population genomics data to identifysignatures of selective sweep have been actively developed, but mostly do notidentify the specific mutation favored by the selective sweep. In this talk, wedescribe an approach that uses population genetics and machine learning tech-niques to pin-point the favored mutation, even when the signature of selectionextends to 5Mbp. Our method, iSAFE, was tested extensively on simulated dataand 22 known sweeps in human populations using the 1000 genome project datawith some evidence for the favored mutation. iSAFE ranked the candidatemutation among the top 15 (out of * 21,000 candidates) in 14 of the 22 loci,and identified previously unreported mutations as favored the 5 regions.
Sequence Recognition as a SubcellularComputational Primitive in Neural Function
Upinder S. Bhalla
National Centre for Biological Sciences (NCBS), Bangalore, [email protected]
Abstract. Many sensory, motor, and cognitive processes involve sequenceswith complex hierarchical structures. In computational neuroscience these havetypically been modeled as arising from network computation. We have analyzedhow such computations may arise instead from subcellular reaction-diffusionprocesses on small (*30 micron) segments of neuronal dendrites. This for-mulation vastly increases the potential computational capacity of neuronalnetworks. We consider some possible mappings of subcellular sequence com-putation to the structure of deep learning networks. This is interesting because itprovides for very compact and efficient biological implementations ofLSTM-like networks. We speculate that there may be a parallel between someof the computational principles of engineered networks and the hippocampal-entorhinal cortex loop.
An Incremental Fast Policy SearchUsing a Single Sample Path
Ajin George Joseph and Shalabh Bhatnagar
Indian Institute of Science, Bangalore, India{ajin,shalabh}@iisc.ac.in
Abstract. In this paper, we consider the control problem in a reinforcementlearning setting with large state and action spaces. The control problem mostcommonly addressed in the contemporary literature is to find an optimal policywhich optimizes the long run c -discounted transition costs, where c 2 0; 1½ Þ .They also assume access to a generative model/simulator of the underlyingMDP with the hidden premise that realization of the system dynamics of theMDP for arbitrary policies in the form of sample paths can be obtained with easefrom the model. In this paper, we consider a cost function which is theexpectation of a approximate value function w.r.t. the steady state distributionof the Markov chain induced by the policy, without having access to the gen-erative model. We assume that a single sample path generated using a priorichosen behaviour policy is made available. In this information restricted setting,we solve the generalized control problem using the incremental cross entropymethod. The proposed algorithm is shown to converge to the solution which isglobally optimal relative to the behaviour policy.
Biometric Counter-Spoofing for MobileDevices Using Gaze Information
Asad Ali , Nawal Alsufyani , Sanaul Hoque ,and Farzin Deravi
School of Engineering and Digital Arts,University of Kent, Canterbury, Kent CT2 7NT, UK
Abstract.With the rise in the use of biometric authentication on mobile devices,it is important to address the security vulnerability of spoofing attacks where anattacker using an artefact representing the biometric features of a genuine userattempts to subvert the system. In this paper, techniques for presentation attackdetection are presented using gaze information with a focus on their applicabilityfor use on mobile devices. Novel features that rely on directing the gaze of theuser and establishing its behaviour are explored for detecting spoofing attempts.The attack scenarios considered in this work include the use of projected photos,2D and 3D masks. The proposed features and the systems based on them wereextensively evaluated using data captured from volunteers performing genuineand spoofing attempts. The results of the evaluations indicate that gaze-basedfeatures have the potential for discriminating between genuine attempts andimposter attacks on mobile devices.
Contents
Invited Talks
An Incremental Fast Policy Search Using a Single Sample Path . . . . . . . . . . 3Ajin George Joseph and Shalabh Bhatnagar
Biometric Counter-Spoofing for Mobile Devices Using Gaze Information . . . 11Asad Ali, Nawal Alsufyani, Sanaul Hoque, and Farzin Deravi
Pattern Recognition and Machine Learning
kNN Classification with an Outlier Informative Distance Measure . . . . . . . . . 21Gautam Bhattacharya, Koushik Ghosh, and Ananda S. Chowdhury
Tree-Based Structural Twin Support Tensor Clusteringwith Square Loss Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Reshma Rastogi and Sweta Sharma
Kernel Entropy Discriminant Analysis for Dimension Reduction . . . . . . . . . . 35Aditya Mehta and C. Chandra Sekhar
A New Method to Address Singularity Problem in MultimodalData Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Ankita Mandal and Pradipta Maji
Label Correlation Propagation for Semi-supervised Multi-label Learning . . . . 52Aritra Ghosh and C. Chandra Sekhar
Formulation of Two Stage Multiple Kernel LearningUsing Regression Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
S.S. Shiju, Asif Salim, and S. Sumitra
A Two-Stage Conditional Random Field Model Based Frameworkfor Multi-Label Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Abhiram Kumar Singh and C. Chandra Sekhar
A Matrix Factorization & Clustering Based Approachfor Transfer Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
V. Sowmini Devi, Vineet Padmanabhan, and Arun K. Pujari
Signal and Image Processing
Feature Selection and Fuzzy Rule Mining for Epileptic Patientsfrom Clinical EEG Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Abhijit Dasgupta, Losiana Nayak, Ritankar Das, Debasis Basu,Preetam Chandra, and Rajat K. De
Selection of Relevant Electrodes Based on Temporal Similarityfor Classification of Motor Imagery Tasks . . . . . . . . . . . . . . . . . . . . . . . . . 96
Jyoti Singh Kirar, Ayesha Choudhary, and R.K. Agrawal
Automated Measurement of Translational Margins and Rotational Shiftsin Pelvic Structures Using CBCT Images of Rectal Cancer Patients. . . . . . . . 103
Sai Phani Kumar Malladi, Bijju Kranthi Veduruparthi,Jayanta Mukherjee, Partha Pratim Das, Saswat Chakrabarti,and Indranil Mallick
Exploring the Scope of HSV Color Channels Towards SimpleShadow Contour Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Jayeeta Saha and Arpitam Chatterjee
Linear Curve Fitting-Based Headline Estimation in HandwrittenWords for Indian Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Rahul Pramanik and Soumen Bag
Object Segmentation in Texture Images Using Texture GradientBased Active Contours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Priyambada Subudhi and Susanta Mukhopadhyay
A Variance Based Image Binarization Scheme and Its Applicationin Text Segmentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Ranjit Ghoshal, Aditya Saha, and Sayan Das
Computer Vision and Video Processing
Variants of Locality Preserving Projection for Modular Faceand Facial Expression Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Gitam Shikkenawis and Suman K. Mitra
A Robust Color Video Watermarking Technique Using DWT, SVDand Frame Difference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Sai Shyam Sharma, Sanik Thapa, and Chaitanya Pavan Tanay
Aggregated Channel Features with Optimum Parametersfor Pedestrian Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Blossom Treesa Bastian and C. Victor Jiji
XXII Contents
Object Tracking with Classification Score Weighted Histogramof Sparse Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Mathew Francis and Prithwijit Guha
A Machine Learning Inspired Approach for Detection, Recognitionand Tracking of Moving Objects from Real-Time Video . . . . . . . . . . . . . . . 170
Anit Chakrabory and Sayandip Dutta
Does Rotation Influence the Estimated Contour Length of a Digital Object? . . . 179Sabyasachi Mukherjee, Oishila Bandyopadhyay, Arindam Biswas,and Bhargab B. Bhattacharya
Abnormal Crowd Behavior Detection Based on CombinedApproach of Energy Model and Threshold . . . . . . . . . . . . . . . . . . . . . . . . . 187
Madhura Halbe, Vibha Vyas, and Yogita M. Vaidya
Unsupervised Feature Descriptors Based Facial Trackingover Distributed Geospatial Subspaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Shubham Dokania, Ayush Chopra, Feroz Ahmad, S. Indu,and Santanu Chaudhury
Face Detection Based on Frequency Domain Features . . . . . . . . . . . . . . . . . 203B.H. Shekar and D.S. Rajesh
A Study on the Properties of 3D Digital Straight Line Segments. . . . . . . . . . 212Mousumi Dutt, Somrita Saha, and Arindam Biswas
Unlocking the Mechanism of Devanagari Letter IdentificationUsing Eye Tracking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Chetan Ralekar, Tapan K. Gandhi, and Santanu Chaudhury
Video Stabilization Using Sliding Frame Window . . . . . . . . . . . . . . . . . . . . 227Keerthan S. Shagrithaya, Eeshwar Gurushankar, Deepak Srikanth,Pravin Bhaskar Ramteke, and Shashidhar G. Koolagudi
Palmprint and Finger Knuckle Based Person Authenticationwith Random Forest via Kernel-2DPCA. . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Gaurav Jaswal, Amit Kaul, and Ravinder Nath
Soft and Natural Computing
A Fuzzy-LP Approach in Time Series Forecasting . . . . . . . . . . . . . . . . . . . 243Pritpal Singh and Gaurav Dhiman
Third Order Backward Elimination Approach for Fuzzy-Rough SetBased Feature Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Soumen Ghosh, P.S.V.S. Sai Prasad, and C. Raghavendra Rao
Contents XXIII
A Novel OCR System Based on Rough Set Semi-reduct . . . . . . . . . . . . . . . 263Ushasi Chaudhuri, Partha Bhowmick, and Jayanta Mukherjee
Rough Set Rules Determine Disease Progressions in Different Groupsof Parkinson’s Patients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Andrzej W. Przybyszewski, Stanislaw Szlufik, Piotr Habela,and Dariusz M. Koziorowski
Adversarial Optimization of Indoor Positioning SystemUsing Differential Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Feroz Ahmad and Sreedevi Indu
Fast Convergence to Near Optimal Solution for Job Shop SchedulingUsing Cat Swarm Optimization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Vivek Dani, Aparna Sarswat, Vishnu Swaroop, Shridhar Domanal,and Ram Mohana Reddy Guddeti
Music-Induced Emotion Classification from the Prefrontal Hemodynamics . . . 289Pallabi Samanta, Diptendu Bhattacharya, Amiyangshu De, Lidia Ghosh,and Amit Konar
Speech and Natural Language Processing
Analysis of Features and Metrics for Alignment in Text-DependentVoice Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Nirmesh J. Shah and Hemant A. Patil
Effectiveness of Mel Scale-Based ESA-IFCC Featuresfor Classification of Natural vs. Spoofed Speech . . . . . . . . . . . . . . . . . . . . . 308
Madhu R. Kamble and Hemant A. Patil
Novel Phase Encoded Mel Filterbank Energies for EnvironmentalSound Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Rishabh N. Tak, Dharmesh M. Agrawal, and Hemant A. Patil
An Adaptive i-Vector Extraction for Speaker Verificationwith Short Utterance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
Arnab Poddar, Md Sahidullah, and Goutam Saha
Spoken Keyword Retrieval Using Source and System Features . . . . . . . . . . . 333Maulik C. Madhavi, Hemant A. Patil, and Nikhil Bhendawade
Novel Gammatone Filterbank Based Spectro-Temporal Featuresfor Robust Phoneme Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
Ankit Nagpal and Hemant A. Patil
XXIV Contents
Neural Networks Compression for Language Modeling . . . . . . . . . . . . . . . . 351Artem M. Grachev, Dmitry I. Ignatov, and Andrey V. Savchenko
A Metaphor Detection Approach Using Cosine Similarity . . . . . . . . . . . . . . 358Malay Pramanick and Pabitra Mitra
Named Entity Identification Based Translation Disambiguation Model . . . . . . 365Vijay Kumar Sharma and Namita Mittal
LEXER: LEXicon Based Emotion AnalyzeR . . . . . . . . . . . . . . . . . . . . . . . 373Shikhar Sharma, Piyush Kumar, and Krishan Kumar
Lexical TF-IDF: An n-gram Feature Space for Cross-Domain Classificationof Sentiment Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
Atanu Dey, Mamata Jenamani, and Jitesh J. Thakkar
A Method for Semantic Relatedness Based Query FocusedText Summarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Nazreena Rahman and Bhogeswar Borah
Bioinformatics and Computational Biology
Efficient and Effective Multiple Protein Sequence Alignment ModelUsing Dynamic Progressive Approach with Novel Look Back AheadScoring System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Sanjay Bankapur and Nagamma Patil
Classification of Vector-Borne Virus Through Totally Ordered Setof Dinucleotide Interval Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
Uddalak Mitra and Balaram Bhattacharyya
A Quasi-Clique Mining Algorithm for Analysis of the HumanProtein-Protein Interaction Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Brijesh Kumar Sriwastava, Subhadip Basu, and Ujjwal Maulik
Prediction of Thyroid Cancer Genes Using an Ensemble of PostTranslational Modification, Semantic and Structural SimilarityBased Clustering Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Anup Kumar Halder, Pritha Dutta, Mahantapas Kundu, Mita Nasipuri,and Subhadip Basu
mRMR+: An Effective Feature Selection Algorithm for Classification . . . . . . 424Hussain A. Chowdhury and Dhruba K. Bhattacharyya
Topological Inquisition into the PPI Networks Associatedwith Human Diseases Through Graphlet Frequency Distribution . . . . . . . . . . 431
Debjani Bhattacharjee, Sk Md Mosaddek Hossain, Raziya Sultana,and Sumanta Ray
Contents XXV
Machine Learning Approach for Identification of miRNA-mRNARegulatory Modules in Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Sushmita Paul and Shubham Talbar
Data Mining and Big Data Analytics
K-Means Algorithm to Identify k1-Most Demanding Products. . . . . . . . . . . . 451Ritesh Kumar, Partha Sarathi Bishnu, and Vandana Bhattacherjee
Detection of Atypical Elements by Transforming Taskto Supervised Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Piotr Kulczycki and Damian Kruszewski
Mining Rare Patterns Using Hyper-Linked Data Structure . . . . . . . . . . . . . . 467Anindita Borah and Bhabesh Nath
Random Binary Search Trees for Approximate Nearest NeighbourSearch in Binary Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
Michał Komorowski and Tomasz Trzciński
A Graphical Model for Football Story Snippet Synthesisfrom Large Scale Commentary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
Anirudh Vyas, Sangram Gaikwad, and Chiranjoy Chattopadhyay
An Efficient Approach for Mining Frequent Subgraphs . . . . . . . . . . . . . . . . 486Tahira Alam, Sabit Anwar Zahin, Md. Samiullah,and Chowdhury Farhan Ahmed
Image Annotation Using Latent Components and Transmedia Association . . . 493Anurag Tripathi, Abhinav Gupta, Santanu Chaudhary, and Brejesh Lall
Incremental Learning of Non-stationary Temporal Causal Networksfor Telecommunication Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Ram Mohan, Santanu Chaudhury, and Brejesh Lall
Effectiveness of Representation and Length Variation of Shortest Pathsin Graph Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
Asif Salim, S.S. Shiju, and S. Sumitra
An Efficient Encoding Scheme for Dynamic Multidimensional Datasets. . . . . 517Mehnuma Tabassum Omar and K.M. Azharul Hasan
Deep Learning
Stacked Features Based CNN for Rotation Invariant Digit Classification . . . . 527Ayushi Jain, Gorthi R.K. Sai Subrahmanyam, and Deepak Mishra
XXVI Contents
Improving the Performance of Deep Learning Based Speech EnhancementSystem Using Fuzzy Restricted Boltzmann Machine . . . . . . . . . . . . . . . . . . 534
Suman Samui, Indrajit Chakrabarti, and Soumya K. Ghosh
A Study on Deep Convolutional Neural Network Based Approachesfor Person Re-identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Harendra Chahar and Neeta Nain
Two-Stream Convolutional Network with Multi-level Feature Fusionfor Categorization of Human Action from Videos . . . . . . . . . . . . . . . . . . . . 549
Prateep Bhattacharjee and Sukhendu Das
Learning Deep Representation for Place Recognition in SLAM. . . . . . . . . . . 557Aritra Mukherjee, Satyaki Chakraborty, and Sanjoy Kumar Saha
Performance of Deep Learning Algorithms vs. Shallow Models, in ExtremeConditions - Some Empirical Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Samik Banerjee, Prateep Bhattacharjee, and Sukhendu Das
Deep Learning in the Domain of Multi-Document Text Summarization . . . . . 575Rajendra Kumar Roul, Jajati Keshari Sahoo, and Rohan Goel
Space-Time Super-Resolution Using Deep Learning Based Framework . . . . . 582Manoj Sharma, Santanu Chaudhury, and Brejesh Lall
A Spatio-temporal Feature Learning Approach for DynamicScene Recognition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Ihsan Ullah and Alfredo Petrosino
Spatial Data Science and Engineering
Spatial Distribution Based Provisional Disease Diagnosisin Remote Healthcare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
Indrani Bhattacharya and Jaya Sil
Extraction of Phenotypic Traits for Drought Stress StudyUsing Hyperspectral Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Swati Bhugra, Nitish Agarwal, Shubham Yadav, Soham Banerjee,Santanu Chaudhury, and Brejesh Lall
Spatio-Temporal Prediction of Meteorological Time Series Data:An Approach Based on Spatial Bayesian Network (SpaBN) . . . . . . . . . . . . . 615
Monidipa Das and Soumya K. Ghosh
Adaptive TerraSAR-X Image Registration (AIR) Using Spatial FisherKernel Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623
B. Sirisha, Chandra Sekhar Paidimarry, A.S. Chandrasekhara Sastry,and B. Sandhya
Contents XXVII
Applications of Pattern Recognition and Machine Intelligence
Hierarchical Ranking of Cricket Teams Incorporating Player Composition . . . 633Abhinav Agarwalla, Madhav Mantri, and Vishal Singh
Smart Water Management: An Ontology-Driven Context-AwareIoT Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
Deepti Goel, Santanu Chaudhury, and Hiranmay Ghosh
Structured Prediction of Music Mood with Twin Gaussian Processes . . . . . . . 647Santosh Chapaneri and Deepak Jayaswal
Differentiating Pen Inks in Handwritten Bank ChequesUsing Multi-layer Perceptron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Prabhat Dansena, Soumen Bag, and Rajarshi Pal
Analysis of Causal Interactions and Predictive Modelling of FinancialMarkets Using Econometric Methods, Maximal Overlap Discrete WaveletTransformation and Machine Learning: A Study in Asian Context . . . . . . . . 664
Indranil Ghosh, Manas K. Sanyal, and R.K. Jana
Opinion Mining Using Support Vector Machine with Web BasedDiverse Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Mir Shahriar Sabuj, Zakia Afrin, and K.M. Azharul Hasan
Harnessing Online News for Sarcasm Detection in Hindi Tweets . . . . . . . . . 679Santosh Kumar Bharti, Korra Sathya Babu, and Sanjay Kumar Jena
Concept-Based Approach for Research Paper Recommendation . . . . . . . . . . 687Ritu Sharma, Dinesh Gopalani, and Yogesh Meena
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
XXVIII Contents