eeecs 2020 - web venue
TRANSCRIPT
1
EEECS 2020 Program and Abstracts
The 7th International Conference on Electronics,
Electrical Engineering, Computer Science 2020
December 21-22, 2020
Gachon University, Korea
Sponsored by
Korea Culture & Contents Technology Association (KOCTA), Immersive Content Display Center (ICDC),
Computer and Communication Engineering for Capacity Building (CCC)
Copyright ⓒ 2020 KOCTA.
Program & Abstracts EEECS2020
2
Contents
1 Messages ........................................................................................................................................... 3
1.1 Message from General Chair ....................................................................................................... 3
2 Committee ........................................................................................................................................ 4
2.1 Organizing Committee ................................................................................................................. 4
2.2 Technical Program Committee ..................................................................................................... 5
3 Keynote Speeches ............................................................................................................................. 7
4 Program at a Glance ........................................................................................................................ 9
5 Technical Program ......................................................................................................................... 11
RS1 Regular Session: Smart Contents Protection ...................................................................... 11
RS2 Regular Session: Multimedia Contents & Systems ............................................................. 11
RS3 Regular Session: Intelligent Signal and Image Processing 1 ............................................... 12
RS5 Regular Session: High Performance Computing Systems ................................................... 12
RS6 Regular Session: Intelligent Signal and Image Processing 2 ............................................... 13
RS7 Regular Session: Immersive Display, Network & Contents ................................................. 13
RS8 Regular Session: Artificial Intelligence for Embedded Systems .......................................... 14
SS1 Special Session: Smart Multimedia Applications ............................................................... 14
SS2 Special Session: Immersive Content Display ...................................................................... 15
6 Abstracts ......................................................................................................................................... 16
RS1 Regular Session: Smart Contents Protection ...................................................................... 16
RS2 Regular Session: Multimedia Contents & Systems ............................................................. 17
RS3 Regular Session: Intelligent Signal and Image Processing 1 ............................................... 19
RS5 Regular Session: High Performance Computing Systems ................................................... 20
RS6 Regular Session: Intelligent Signal and Image Processing 2 ............................................... 22
RS7 Regular Session: Immersive Display, Network & Contents ................................................. 23
RS8 Regular Session: Artificial Intelligence for Embedded Systems .......................................... 24
SS1 Special Session: Smart Multimedia Applications ............................................................... 26
SS2 Special Session: Immersive Content Display ...................................................................... 27
7 Author Index .................................................................................................................................. 30
EEECS2020 Program & Abstracts
3
1 Messages
1.1 Message from General Chair
It is a great pleasure for me to welcome you to the 7th International Conference on
Electronics, Electrical Engineering, Computer Science (7th EEECS 2017) from December
21 to December 22, 2020. This year’s conference marks the seventh EEECS starting from
2016. The EEECS is a conference of the Korea Culture & Contents Technology Association
(KOCTA) and represents large number of gatherings of researchers and industry
professionals in the corresponding fields.
This year’s conference brings together more than 50 delegates from around the Asian
countries to discuss the latest advances in this vibrant and constantly evolving field. The
topics covered in the program include overall areas in Electronics, Electrical Engineering,
and Computer Science. In line with recent research trends, many artificial intelligence-related
papers have been accepted with the diligent work of the technical program committee.
2020 has been a very challenging year due to the COVID-19 pandemic. This conference
is originally planned to be held in Gachon University, Korea. But due to the COVID-19
spread out, it has been converted to a fully online conference. Nonetheless this difficulty
situation, the committee would thank all participants and paper authors contributing this
conference more active. Through this online platform, EEECS 2020 continues to share an
insight into the recent research and cutting-edge technologies in those fields of ICT.
The EEECS2020 has been made up by many volunteers who contributed to the various
processes and it would not be possible for me to name all of them in this short message. In
particular, the Technical Program Committee, led by our indefatigable TPC Chairs and
supported by the TPC members, completed a thorough peer-review process of technical and
special session papers to select a comprehensive and high-quality technical program for the
conference. This program is augmented and complemented by two Keynote Speeches, two
special sessions and several regular sessions. In addition, all Organizing Committees worked
tirelessly to ensure the best quality experience for the delegates during the technical sessions
and the social programs.
Also I would like to thank the groups of KOCTA and ICDC, Kwangwoon University,
Korea and CCC, Mae Fah Luang University. Finally, I would also like to thank all participants
and supporters for their contribution to the conference. It is a fantastic experience for me to
serve as the General Chair of EEECS2020 and it is my hope that you find the conference
stimulating, fulfilling and enjoyable. Please enjoy the conference!
7th EEECS 2020 General Chair
Taek Keun Hwangbo
Program & Abstracts EEECS2020
4
2 Committee
2.1 Organizing Committee
Honorary Chair
Byunggi Kim, Soongsil University, Korea
Kosin Chamnongthai, KMUTT, Thaniland
General Chair
Taek-Geun HwangBo, Gachon University, Korea
Technical Program Co-Chairs
Seung Hyun Lee, Kwangwoon University, Korea
Kyoungro Yoon, Konkuk University, Korea
Chayapol Kamyod, Mae Fah Luang University, Thailand
Special Session Chairs
Jiman Hong, Soongsil University, Korea
Sang Kyun Kim, Myungji University, Korea
Sangwoon Lee, Namseoul University, Korea
Workshop Chairs
Jin Young Kim, Kwangwoon University, Korea
Youngseop Kim, Dankook University, Korea
Publicity Chairs
Young-Ho Seo, Kwangwoon University, Korea
Hae Chul Choi, Hanbat National University, Korea
Dong Myung Shin, LSWare Inc, Korea
Publication Chairs
Youngmo Kim, Soongsil University, Korea
Finance & Registration Co-Chairs
Youngwhan Lee, Wonkwang University, Korea
Cheong Ghil Kim, Namseoul University, Korea
Local Arrangement Chair
Seok Hee Oh, Gachon University, Korea
Information System Chair
Seungmin Lee, Namseoul University, Korea
Ui Jin Jang, Soongsil University, Korea
General Secretaries
Su-Kyung Yoon, Jeonbuk National University, Korea
EEECS2020 Program & Abstracts
5
2.2 Technical Program Committee
Chair
Seung Hyun Lee, Kwangwoon University, Korea
Kyoungro Yoon, Konkuk University, Korea
Chayapol Kamyod, Mae Fah Luang University, Thailand
Members
Jiman Hong, Soongsil University, Korea
Hidehiro Kanemitsu, Tokyo University of Technology, Japan
Jin Young Kim, Kwangwoon University
Young-Ho Seo, Kwangwoon University, Korea
Kosin Chamnongthai, KMUTT, Thaniland
Punnarumol Temdee, Mae Fah Luang University, Thailand
Hamed Yahoui, Université Lyon 1, France
Youngmo Kim, Soongsil University, Korea
Roungsan Chaisricharoen, Mae Fah Luang University, Thailand
Jun-yu Dong, Ocean University of China, China
Muwei Jian, Shandong University of Finance and Economics, China
Nattapol Aunsri, Mae Fah Luang University, Thailand
Seok Hee Oh, Gachon University, Korea
Santichao Wicha, Mae Fah Luang University, Thailand
Sang Kyun Kim, Myungji University, Korea
Su-Kyung Yoon, Jeonbuk National University, Korea
Seok Yoon Kim, Soongsil University, Korea
Ji Hwan Kim, Sogang University, Korea
Hae Chul Choi, Hanbat National University, Korea
Ui Jin Jang, Soongsil University, Korea
Moo Wan Kim, Tokyo University of Information Sciences, Japan
Youngseop Kim, Dankook University, Korea
Tae Young Byun, Daegu Catholic University
Su-Yeon Kim, Deagu University
Apiradee Ampawasiri, Provincial Electricity Authority, Thailand
Cheong Ghil Kim, Namseoul University, Korea
Choong Pyo Hong, LG Electronics, Korea
Chompoo Suppatoomsin, Vongchavalitkul University, Thailand
Byung In Moon, Kyungpook National University, Korea
Fumitaka Ono, Tokyo Polytechnic University, Japan
Guodong Wang, Qingdao University, China
Hae Kyung Chung, Konkuk University, Korea
Program & Abstracts EEECS2020
6
Hui Xia, Qingdao University, China
Jae-sang Cha, Seoul National University of Science and Technology, Korea
Je Ho Park, Dankook University, Korea
Jia Zhao, Nanchang Institute of Technology, China
Jianbo Li, Qingdao University, China
Jin Ho Ahn, Hoseo University, Korea
Jung Hoon Lee, Gyeongsang National University, Korea
Seungmin Lee, Namseoul University, Korea
Sunghwa Lim, Namseoul University, Korea
Muhammad Arshad Awan, Allama Iqbal Open University, Pakistan
Qian Zhang, Taishan University, China
Sang Woon Lee, Namseoul University, Korea
Sasalak Tongkaw, Songkhla Rajabhat University, Thailand
Sooncheol Kwon, Kwangwoon University, Korea
Takaaki Ishikawa, Waseda University, Japan
Won Gee Hong, Daegu University, Korea
Woo Chan Park, Sejong University, Korea
Taebum Lim, KETI, Korea
Yiying Zhang, Tianjin University of Science&Technology, China
Yong Hwan Lee, Wonkwang University, Korea
Yongsoo Choi, Sungkyul University, Korea
Young Choong Park, KETI, Korea
Young Ho Seo, Mokwon University, Korea
EEECS2020 Program & Abstracts
7
3 Keynote Speeches
Monday, 21st December 2020, 14:30 - 15:30
Categorization and Measurement by using heat information
Prof. Kosin Chamnongthai
King Mongkut's University of Technology Thonburi
Abstract
Thermal signal and thermal image recently become another important and interesting information
comparing with conventional signals representingbrightness, sound, distance, etc. In some
conditions such as nighttime, it is almost impossible to get brightness information to process and
categorize object types. As a case study, thermal signal is regarded to be able to apply for vehicle-
type categorizationat night time and dark areas, since vehicles generally gets heat from their
engines, and their heat patterns can be categorized and recognized the vehicle types. Similarly, food
that is normally cooked in the warm or even hot temperature level is regardedas possible to classify
and recognize food types and their ingredients by using heat pattern. In this talk, research cases of
nighttime vehicle-type categorization and food calorie measurement are explained and discussed in
order to understand the trend ofthermal signal applications.
Biography
Kosin Chamnongthai currently works as professor at
Electronic and Telecommunication Engineering Department,
Faculty of Engineering, King Mongkut's University of
Technology Thonburi (KMUTT), and also serves as
presidentof ECTI Association (2018-2019). He served as
editor of ECTI e-magazine during 2011-2015, associate
editor of ECTI-CIT Trans during 2011-2016, associate editor
of ECTI-EEC Trans during 2003-2010, associate editor of
ELEX (IEICE Trans) during 2008-2010, andchairman of
IEEE COMSOC Thailand during 2004-2007.
He has received B.Eng. in Applied Electronics from the University of Electro-communications,
Tokyo, Japan in 1985, M.Eng. in Electrical Engineering from Nippon Institute of Technology, Saitama,
Japan in 1987, and Ph.D.in Electrical Engineering from Keio University, Tokyo, Japan in 1991. His
research interests include computer vision, image processing, robot vision, signal processing, and
pattern recognition. He is a senior member of IEEE, and a member of IEICE, TESA, ECTI,AIAT,
APSIPA, TRS, and EEAAT.
Program & Abstracts EEECS2020
8
Monday, 21st December 2020, 13:40 - 14:30
Toward autonomous container-based task scheduling for efficient IoT
processing
Prof. Hidehiro Kanemitsu
Tokyo University of Technology
Abstract
Current information processing pardaigms handle IoT data, multimedia streams, large volume
file, and so on. Objectives of such information processing involeves data analysis, data format
transformation, and calcuration. In IoT systems, various kinds of data should be processed
efficiently in heterogeneous systems across regions to share information sharing, e.g., among smart
cities. As for the processing system, a virtualized environment such as cloud and container-based
ones have been adopted for utilize application processes among computatinal resources. Thus, one
of current and future issues in terms of various data processing models is how each "task" should be
processed across heterogeneous virtualized systems. In this presentation, I introduce our research
topics for container-based task allocation and scheduling schemes on multiple clouds. The topic
includes algorithms for chaining each container-based service function (SF) in order to process IoT
data efficiently.
Biography
Hidehiro Kanemitsu received his B.S. degreein Science from Waseda
University, Japan, andM.S. and Ph.D. degrees in Global Informationand
Telecommunication Studies from WasedaUniversity, Japan. His research
interests includeparallel and distributed computing, grids, peerto-peer
computing, and web service technology. He has published several
prestigious journals and proceedings in terms of distributed computing
area such as IEEE TPDS, JPDC, IEEE CLOUD, and IEEE ICWS. He is
currently a senior assistant professor at theSchool of Computer Science,
Tokyo Universityof Technology, Japan. He is a member of IEICE,IPSJ,
and IEEE.
EEECS2020 Program & Abstracts
9
4 Program at a Glance
Monday, 21st December 2020
10:45 – 12:00 Regular Session 1: Smart Contents Protection
Chair: Su-Kyung Yoon (Jeonbuk National University, Korea)
Papers: RS1-1, RS1-2, RS1-3, RS1-4, RS1-5
12:00 – 13:15 Lunch Break
13:15 – 13:30 Pre-arrangement
Plenary Session:
Chair: Chayapol Kamyod (Mae Fah Luang University, Thailand)
Message from General Chair:
Taek-Geun HwangBo (Gachon University, Korea)
Keynote Speech: Toward autonomous container-based task
scheduling for efficient IoT processing
Invited Speaker: Prof. Hidehiro Kanemitsu
(Tokyo University of Technology, Japan)
Keynote Speech: Categorization and Measurement by using heat
information
Invited Speaker: Prof. Kosin Chamnongthai
(King Mongkut's University of Technology Thonburi,
Thailand)
13:30 – 13:40
13:40 – 14:30
14:30 – 15:30
15:30 – 15:45 Break
15:45 – 17:00 Special Session 1: Smart Multimedia Applications
Chair: Jeong-geun Kim (Yonsei University, Korea)
Papers: SS1-1, SS1-2, SS1-3, SS1-4, SS1-5
Special Session 2: Immersive Content Display
Chair: Sunghwa Lim (Namseoul University, Korea)
Papers: SS2-1, SS2-2, SS2-3, SS2-4, SS2-5
Regular Session 8: Artificial Intelligence for Embedded Systems
Chair: Youngmo Kim (Soongsil University, Korea)
Papers: RS8-1, RS8-2, RS8-3, RS8-4, RS8-5
Program & Abstracts EEECS2020
10
Tuesday, 22nd December 2020
10:45 – 12:00 Regular Session 2: Multimedia Contents & Systems
Chair: Seok Hee Oh (Gachon University, Korea)
Papers: RS2-1, RS2-2, RS2-3, RS2-4, RS2-5
Regular Session 5: High Performance Computing Systems
Chair: Yong Hwan Lee (Wonkwang University, Korea)
Papers: RS5-1, RS5-2, RS5-3, RS5-4, RS5-5
12:00 – 13:15 Lunch Break
13:15 – 14:30 Regular Session 3: Intelligent Signal and Image Processing 1
Chair: Youngseop Kim (Dankook University, Korea)
Papers: RS3-1, RS3-2, RS3-3, RS3-4, RS3-5
Regular Session 7: Immersive Display, Network & Contents
Chair: Jin Young Kim (Kwangwoon University, Korea)
Papers: RS7-1, RS7-2, RS7-3, RS7-4
14:30 – 14:45 Break
14:45 – 16:00 Regular Session 6: Intelligent Signal and Image Processing 2
Chair: Kyoungro Yoon (Konkuk University, Korea)
Papers: RS6-1, RS6-2, RS6-3, RS6-4, RS6-5
16:00 – 16:10 Break
16:10 – 17:00 Conference Closing and Award
EEECS2020 Program & Abstracts
11
5 Technical Program
RS1 Regular Session: Smart Contents Protection
Monday, 21st December 2020, 10:45 – 12:00
Chair: Su-Kyung Yoon (Jeonbuk National University, Korea)
RS1-1
Cached Transaction Executor for Distributed Computing of Single Ledger
Blockchain
EEECS204
YongJoon Joe, Seong-Bo Kim, Dong-Myong Shin (LSware Inc.)
RS1-2
An Original Video Detection Method Using Feature Information of
Immersive 360-Degree Video
EEECS228
Injae Yoo, Jaecheng Lee (Beyond Tech Inc.), Byeongchan Park, Youngmo Kim, Seok-
Yoon Kim (Soongsil University)
RS1-3
Metadata Structure of Usage History for Transparent Settlement and
Distribution of Theme, Background and Signal Music
EEECS229
Ulugbek Ruizive, Byeongchan Park, Youngmo Kim, Seok-Yoon Kim (Soongsil
University)
RS1-4
A Study on Usage Duplication Avoiding Method Using Blockchain in
Theme, Background and Signal Sound Sources
EEECS231
Seyoung Jang, Seok-Yoon Kim (Soongsil University)
RS1-5
A Study on Performance Enhancement of BI-LSTM Single-Text
Sentiment Model Using Bootstrap Method with Limitation in Number of
Tokens
EEECS232
Seunghyun Ji, Youngmo Kim, Seok-Yoon Kim (Soongsil University)
RS2 Regular Session: Multimedia Contents & Systems
Tuesday, 22nd December 2020, 10:45 – 12:00
Chair: Seok Hee Oh (Gachon University, Korea)
RS2-1
Energy-efficient Heterogenous Multiple Wireless Interfaces for
Smartphones
EEECS239
Se Won Lee (Pukyong National University), Sung-Hwa Lim (Namseoul University)
RS2-2
Automatic Vending Device Capable of Ordering Products Through Voice
and Control Method Thereof
EEECS207
HaeKyung Chung (Konkuk University), JangHyok Ko (Shamyook University)
RS2-3
User Experience Study of Unmanned Order Payment Kiosk in Fast Food
Store
EEECS216
Seungmin Lee (Namseoul University)
Program & Abstracts EEECS2020
12
RS2-4 Apparatus and Method that Guides the Information of The Locker EEECS215
JangHyok Ko (Shamyook University), HaeKyung Chung (Konkuk University)
RS2-5 Finding Longest Path on a Task Graph for Smartphone Applications EEECS244
Se Won Lee (Pukyong National University), Sung-Hwa Lim (Namseoul University)
RS3 Regular Session: Intelligent Signal and Image Processing 1
Tuesday, 22nd December 2020, 13:15 – 14:30
Chair: Youngseop Kim (Dankook University, Korea)
RS3-1 Optimized Image Link of Audio Video Bridge (AVB) System EEECS208
Byoungman An, Youngseop Kim (Dankook University)
RS3-2 Vehicle Tracking based on Deep Learning EEECS218
Hyochang Ahn (Far East University), Yong-Hwan Lee (Wonkwang University)
RS3-3 Video-based Point Cloud Compression with Versatile Video Coding EEECS233
Daehyeok Kwon, Haechul Choi (Hanbat National University)
RS3-4
Enhancement of 3d Point Cloud Data Using Orthogonal Projection and
Super Resolution Network
EEECS248
Seonghwan Park, Seongbae Rhee, Kyuheon Kim, Hyukmin Kwon, Jeongil Seo (Kyung
Hee University)
RS3-5
Performance Analysis of Subword Tokenization Methods for Language
Modeling
EEECS241
Hosung Park, Hyunsoon Son, Gyujin Kim, Ji-Hwan Kim (Sogang University)
RS5 Regular Session: High Performance Computing Systems
Tuesday, 22nd December 2020, 10:45 – 12:00
Chair: Yong Hwan Lee (Wonkwang University, Korea)
RS5-1 Bank and buffer system design for low power L1 cache memory EEECS217
Bosung Jung, Junghoon Lee (Gyeongsang National University)
RS5-2 PCM Buffer Cache on PCM Hybrid Memory for IoT Devices EEECS210
Su-Kyung Yoon (Jeonbuk National University)
RS5-3 Data Management Techniques for Graph Processing EEECS211
Su-Kyung Yoon (Jeonbuk National University)
RS5-4
Implementation of KD-tree Builder for Real-time Ray and Sound Tracing
Applications
EEECS225
EEECS2020 Program & Abstracts
13
Woonam Chung, Sukwon Choi, Woochan Park (Sejong University), Cheong-Ghil Kim
(Namseoul University)
RS5-5 A Design of Inverse Cosine Calculator for Sound Tracing Applications EEECS226
JinYoung Lee, Woochan Park (Sejong University), Cheong-Ghil Kim (Namseoul
University)
RS6 Regular Session: Intelligent Signal and Image Processing 2
Tuesday, 22nd December 2020, 14:45 – 16:00
Chair: Kyoungro Yoon (Konkuk University, Korea)
RS6-1 Evaluation on Single Image Reflection Removal EEECS219
Hyochang Ahn (Far East University), Yong-Hwan Lee (Wonkwang University)
RS6-2
Deep Learning Based Viewport Super Resolution by Using 2d Warping
Method for 360-Degree Video
EEECS249
Seongbae Rhee, Seonghwan Park, Kyuheon Kim (Kyung Hee University)
RS6-3 Subjective quality evaluation of point clouds compared to natural video EEECS234
Aram Baek, Haechul Choi (Hanbat National University)
RS6-4
Frequency-Tuned Spectrogram with Residual Convolutional Neural
Networks for Acoustic Scene Classification in Muti-Device Environment
EEECS242
Soonshin Seo, Changmin Kim, Donghyun Lee, Hosung Park, Hyunsoon Son, Gyujin
Kim, Ji-Hwan Kim (Sogang University)
RS6-5 Comparison of CNN and YOLO for Object Detection EEECS220
Yong-Hwan Lee (Wonkwang University), Youngseop Kim (Dankook University)
RS7 Regular Session: Immersive Display, Network & Contents
Tuesday, 22nd December 2020, 13:15 – 14:30
Chair: Jin Young Kim (Kwangwoon University, Korea)
RS7-1 ANN-based Body Temperature Monitoring for Home Healthcare System EEECS252
Kyu-Nam Choi, Taeg-Keun Hwangbo (Gachon University)
RS7-2 Promoting of Online PBL For Collaborative Skills Enhancement EEECS250
Santichai Wicha, Panarumol Temdee (Mae Fah Luang University)
RS7-3 CIA And IoT Communication Models EEECS251
Chayapol Kamyod (Mae Fah Luang University)
Program & Abstracts EEECS2020
14
RS7-4
A Study of Proxy Based Information Crawling on Multi-Screen Digital
Signage
EEECS237
Heechan Yang, Daeseung Park, Sangwoon Lee, Cheong Ghil Kim (Namseoul
University), Chayapol Kamyode (Mae Fah Luang University)
RS8 Regular Session: Artificial Intelligence for Embedded Systems
Monday, 21st December 2020, 15:45 – 17:00
Chair: Youngmo Kim (Soongsil University, Korea)
RS8-1
Implementation of Contactless Vehicular Ambient Lighting Control
System Based on the Machine Learning Method
EEECS206
SangYub Lee, Youngchan Kim (Korea Electronics Technology Institute)
RS8-2
Development of KNN-based Haar face detection system using low-power
neuromorphic chip
EEECS212
YoungChan Kim, YeoWool Lee, Inpyo Cho, SangYub Lee (Korea Electronics
Technology Institute)
RS8-3
Uninserted Error Checking Technique Using RBF Tester for Low Power
Low Cost PCB Tester
EEECS213
InPyo Cho, YeoWool Lee, JaeKyu Lee, SangYub Lee (Korea Electronics Technology
Institute)
RS8-4
A Study on the Structure of Learning Image Dataset for Object
Recognition in Construction Sites based on Transfer Learning
EEECS214
Jaekyu Lee, Chulgoo Kim, InPyo Cho, Sangyub Lee (Korea Electronics Technology
Institute)
RS8-5
Development of a digital twin sensor platform for disaster monitoring and
management
EEECS221
Changsuk Yoon, Minsang Yu, Young-Han Kim, Hyun-Seok Ahn, Tae-Beom Lim
(Korea Electronics Technology Institute)
SS1 Special Session: Smart Multimedia Applications
Monday, 21st December 2020, 15:45 – 17:00
Chair: Jeong-geun Kim (Yonsei University, Korea)
SS1-1
A Contactless Personal Health Monitoring System Using Raspberry Pi
Based Smart Mirror
EEECS236
Jiwoong Han, Changhoon Kim, Daeseung Park, Sangwoon Lee, Cheong Ghil Kim
(Namseoul University)
SS1-2
Problem-Based Learning Model for Students Knowledge Acquisition of
Logical Design for Interactive Programming Course
EEECS240
EEECS2020 Program & Abstracts
15
Waralak Chongdarakul, Khwunta Kirimasthong (Mae Fah Luang University)
SS1-3
An Assistive Technology for the Elderly Farmers: A Case Study of a
Cymbidium Orchid Caring System
EEECS245
Kesmanee Runjarean, Nattanicha Thammakhankhom, Warisara Luetan, Prasara
Jakkaew, Sirichai Hemrungrote (Mae Fah Luang University)
SS1-4
The Combined Model of Blended Learning with Flipped Classroom: A
New Learning Strategy for Covid-19 Pandemic Situation
EEECS246
Worasak Rueangsirarak, Punnarumol Temdee (Mae Fah Luang University)
SS1-5
Applying Problem-Based Learning with Direct Instruction Improve
Students Critical Thinking and Teamwork in Thai Undergraduates
EEECS247
Nilubon Kurubanjerdjit, Soontarin Nupap (Mae Fah Luang University)
SS2 Special Session: Immersive Content Display
Monday, 21st December 2020, 15:45 – 17:00
Chair: Sunghwa Lim (Namseoul University, Korea)
SS2-1
DUET-Based Separation of Multiple User Signals for Digital Drone
Signage Systems
EEECS205
Isaac Sim, Young Ghyu Sun, Soo Hyun Kim, Donggu Lee, Jiyoung Lee, Jin Young Kim
(Kwangwoon University)
SS2-2 A Study on The Performance Evaluation of Webxr Device Api EEECS223
Daehyeon Lee, Munyoung Lee, Kye-Dong Jung, Soonchul Kwon (Kwangwoon
University)
SS2-3 A Study on Digital Signage Data Transmission System using RDS EEECS224
SangWoon Lee (Namseoul University)
SS2-4
A Study of Smart Contents Management System Using Machine
Learning for Digital Signage
EEECS235
Daeseung Park, Jiwoong Han, Sangwoon Lee, Cheong Ghil Kim (Namseoul
University)
SS2-5
System Modeling and Implementation of XR Environment -Focused on
VFX Production for TV Drama-
EEECS222
Jungwoon Park, Seokhee Oh (Gachon University)
Program & Abstracts EEECS2020
16
6 Abstracts
RS1 Regular Session: Smart Contents Protection
Monday, 21st December 2020, 10:45 – 12:00
Chair: Su-Kyung Yoon (Jeonbuk National University, Korea)
RS1-
1
Cached Transaction Executor for Distributed Computing of Single Ledger
Blockchain
EEECS204
YongJoon Joe LSware Inc.
Seong-Bo Kim LSware Inc.
Dong-Myong Shin LSware Inc.
Efficient distributed and concurrent computing for blockchain is a complex problem, especially when achieving
scalability with multiple nodes. We approach this problem by preprocessed computing plan by RWKS(Read/Write Key
Set information) pre-checking. This paper shows a value preloading design which utilizes RWKS; a by-product of the
distributed computing planning. Based on this feature, we could accelerate the transaction execution speed by relaxing
the I/O load.
RS1-2
An Original Video Detection Method Using Feature Information of
Immersive 360-Degree Video
EEECS228
Injae Yoo Beyond Tech Inc.
Jaecheng Lee Beyond Tech Inc.
Byeongchan Park Soongsil University
Youngmo Kim Soongsil University
Seok-Yoon Kim Soongsil University
As the immersive 360-degree videos are distributed, piracy activity is also increasing, causing copyright
infringement. Filtering technology to determine the illegal reproduction has been mainly researched and used mainly for
the 2D videos, and it is difficult to apply that technology to immersive 360-degree videos without modification, so a new
filtering technology for them is required. In this paper, we propose an original video detection method using feature
information to determine whether or not a immersive 360-degree video is illegally copied. The proposed method detects
the original video by extracting robust feature information such as size change, brightness change, resolution change, and
rotation change of the immersive 360-degree video, and comparing it with the immersive 360-degree video to which the
attack was applied.
RS1-3
Metadata Structure of Usage History for Transparent Settlement and
Distribution of Theme, Background and Signal Music
EEECS229
Ulugbek Ruizive Soongsil University
Byeongchan Park Soongsil University
Youngmo Kim Soongsil University
Seok-Yoon Kim Soongsil University
It is difficult to transparently settle and distribute the theme, background and signal music, of which the monitoring
is difficult because it is usually used outside of common music. In addition, it is hard to grasp the exact usage history due
EEECS2020 Program & Abstracts
17
to different rights management information among each related organizations. In this paper, we propose a metadata
structure to integrate different music rights management informations and to write usage history information on the
blockchain ledger in the form of smart contract, which can be used as the usage monitoring information for transparent
settlement and distribution of theme, background and signal music.
RS1-4
A Study on Usage Duplication Avoiding Method Using Blockchain in Theme,
Background and Signal Sound Sources
EEECS231
Seyoung Jang Soongsil University
Seok-Yoon Kim Soongsil University
The current process of collecting usage fees in theme, backgrounds and signal(TMS) sound sources other than
common music are mainly categorized into two types, flat-rate and usage-based system used according to the license
compensation method and usage permission types. However, the duplicated/conflicted usage information often causes
credibility issues, which makes the transparent settlement and distribution process of TMS sound sources difficult. To
improve the credibility of TMS sound source usage information, in this paper, we propose a method to prevent usage
duplication/conflict of sound sources using a blockchain technology that processes usage information.
RS1-5
A Study on Performance Enhancement of BI-LSTM Single-Text Sentiment
Model Using Bootstrap Method with Limitation in Number of Tokens
EEECS232
Seunghyun Ji Soongsil University
Youngmo Kim Soongsil University
Seok-Yoon Kim Soongsil University
Since the performance of Bi-LSTM model for single-text sentiment analysis is affected by the volume of training
dataset and complexity of the model, the performance of prediction may be degraded if the size of the dataset is small or
the number of tokens gets bigger. To relieve this problem, this paper proposes a performance enhancement method that
uses bootstrap technique, token normalization and limitation in number of tokens. The experimental results of the
proposed method shows that it has led to 5% higher accuracy than the previous sentiment analysis method.
RS2 Regular Session: Multimedia Contents & Systems
Tuesday, 22nd December 2020, 10:45 – 12:00
Chair: Seok Hee Oh (Gachon University, Korea)
RS2-1 Energy-efficient Heterogenous Multiple Wireless Interfaces for Smartphones EEECS239
Se Won Lee Pukyong National University
Sung-Hwa Lim Namseoul University
The wireless communication module is one of the devices that consume the most energy of a smart mobile terminal.
If we can implement an multi-channel multi-interface environment using these heterogeneous communication modules,
data transmission performance can be improved by increasing parallelism. Also, these heterogeneous modules have
different data rates, transmission ranges, and power consumption. In this paper, we propose a power efficient data
transmission method using heterogeneous communication networks.
RS2-2
Automatic Vending Device Capable of Ordering Products Through Voice and
Control Method Thereof
EEECS207
HaeKyung Chung Konkuk University
Program & Abstracts EEECS2020
18
JangHyok Ko Shamyook University
Disclosed is an automatic sales device capable of ordering products through voice and a control method thereof, and
the automatic sales device capable of ordering products through voice according to an embodiment of the present
application accommodates a plurality of products, and based on product selection information A product receiving unit
in communication with the product discharge port through which the selected product is discharged, a product information
display unit displaying the plurality of products so that the user can identify them, a voice receiving the user's voice input
and extracting the product selection information from the voice input It may include a recognition unit and a control unit
for controlling the product selected based on the product selection information to move to the product discharge port.
RS2-3
User Experience Study of Unmanned Order Payment Kiosk in Fast Food
Store
EEECS216
Seungmin Lee Namseoul University
Recently, domestic and foreign studies on the unmanned order payment system of the restaurant industry, which are
rapidly growing recently, are insufficient. Therefore, there is a lack of understanding of users' perception of unmanned
order settlement system. The purpose of this study is to investigate the effect of user experience factors of unmanned
order payment service of fast food restaurant on satisfaction The purpose of this study is to examine the relationship
between user's satisfaction and intention. Through this study, we can understand the psychology and behavior of users
who use unmanned order settlement service and prepare a strategic foundation that can be used in various industries.
RS2-4 Apparatus and Method that Guides the Information of The Locker EEECS215
JangHyok Ko Shamyook University
HaeKyung Chung Konkuk University
The goods storage application can acquire information about the size of the goods that the user wants to store in
advance, and obtain information on the usage status of the goods storage box (information on the presence or absence of
empty space, etc.). It is possible to effectively provide only information on a storage box having a slot that enables the
actual storage of goods (owned goods). In addition, it provides the function to reserve the slot of the inventory box in
advance, and can guide the exact location of the reserved inventory box, so that users can quickly find the reserved
inventory box without hesitation. In particular, by providing AR-based directions information (internal directions
information), it is possible to help you find the location of the storage box you want to visit faster and at a glance.This
service can provide service functions such as extended payment and reservation function. In addition, it can be easily
used not only for domestic users but also for overseas users, so you can expect tourism effects. For example, an item for
selecting a language, a sign-up window display item, and the like may be displayed on the main screen along with an
intuitive logo. Therefore, not only domestic users but also foreign users (foreign users) can simply subscribe and easily
receive the services provided. The application can determine the user's current location through a GPS sensor and provide
accurate information about the matching item storage box based on this. In addition, it is possible to provide a service
function that allows direct guidance of an optimal place to store items through destination search.
RS2-5 Finding Longest Path on a Task Graph for Smartphone Applications EEECS244
Se Won Lee Pukyong National University
Sung-Hwa Lim Namseoul University
In a smartphone application, a bunch of tasks should be executed in serial or in parallel with some dependency to
complete a mission. Therefore, the task group to be executed can be a directed acyclic graph (i.e., task graph). For real-
time applications, it is very important to find out the expected time to complete the given task graph to see whether the
task graph will finished within its deadline. In this paper, we studied a method to find out the longest path on a task graph
for real-time smartphone applications using PERT/CPM approach.
EEECS2020 Program & Abstracts
19
RS3 Regular Session: Intelligent Signal and Image Processing 1
Tuesday, 22nd December 2020, 13:15 – 14:30
Chair: Youngseop Kim (Dankook University, Korea)
RS3-1 Optimized Image Link of Audio Video Bridge (AVB) System EEECS208
Byoungman An Dankook University
Youngseop Kim Dankook University
Recently, processing huge amount of video data in vehicle network has become a very important task. It is significant
not only for a transmission management but also, for reducing a latency between devices. In this paper, we propose a new
design and optimization method of Audio Video Bridge (AVB) technology. A proposal plays a significant role in reducing
the delay between a talker device and a listener device. The approach on realistic test cases showed that there was a delay
reduction about 52%. It is expected that the optimization method for in-vehicle environment can greatly shorten the time
period in the design and development process. It will greatly benefit the industry since analyzing the latency between
each function in a short period of time is very meaningful.
RS3-2 Vehicle Tracking based on Deep Learning EEECS218
Hyochang Ahn Far East University
Yong-Hwan Lee Wonkwang University
Since traffic jams and accidents are becoming a big social problem, vehicle detecting and tracking are useful to
provide important information such as identifying road traffic conditions. However, vehicle detecting and tracking based
on camera image are affected by environmental factors, such as camera installation and illumination. This paper proposes
a deep learning-based vehicle segmentation and tracking method to classify and track vehicles in a complex background.
Using YOLO model as deep learning, the proposed scheme is more effective and robust to track vehicles in various
environments.
RS3-3 Video-based Point Cloud Compression with Versatile Video Coding EEECS233
Daehyeok Kwon Hanbat National University
Haechul Choi Hanbat National University
A point cloud is a 3D data representation used in immersive media applications including virtual/augmented reality,
immersive telepresence, autonomous driving and cultural heritage archival. Recently, the Motion Picture Experts Group,
a international standard body, has released video-based point cloud compression (V-PCC) to represent and compress point
cloud data. The V-PCC converts the point cloud data from 3D to 2D and then the converted 2D data is coded by a legacy
video encoder. The refernce software of V-PCC, called Test Model category 2, emploies the High Efficiency Video Coding
(HEVC) standard as the video encoder. Recently, Versatiel Video Coding (VCC) standard is released by the MPEG, which
has better coding efficiency than HEVC. This paper introduce a V-PCC based coding method that replaces HEVC with
VVC to compress point cloud data higher. Some VVC coding tools are selectively turned on or off accoridng to their
coding performance for point cloud data. Experimental results show that the proposed scheme achieves average of 23.7%
BD-rate reduction in D1, 32.6% BD-rate reduction in D2, and 22.1% BD-rate reduction in luma component, respectively.
RS3-4
Enhancement of 3d Point Cloud Data Using Orthogonal Projection and Super
Resolution Network
EEECS248
Seonghwan Park Kyung Hee University
Seongbae Rhee Kyung Hee University
Program & Abstracts EEECS2020
20
Kyuheon Kim Kyung Hee University
Hyukmin Kwon Kyung Hee University
Jeongil Seo Kyung Hee University
With the development of media technology, research to provide more immersive media is actively progressing.
Especially, the technology on improving image quality such as super resolution by using deep learning network is widely
researched. Also, 3D point cloud is one of the emerging immersive media contents, which represents 3D media using 3
dimensional coordinates with color information. In this paper, we propose a method of enhancing 3D point cloud data by
applying deep learning network to 3D point cloud data with orthogonal projection.
RS3-5
Performance Analysis of Subword Tokenization Methods for Language
Modeling
EEECS241
Hosung Park Sogang University
Hyunsoon Son Sogang University
Gyujin Kim Sogang University
Ji-Hwan Kim Sogang University
In this paper, unigram-based subword tokenization is applied to language models. Subword tokenization is an
effective way to avoid Out-of-Vocabulary (OOV) problems in speech recognition, natural language processing (NLP).
These language-related processings require a vocabulary generated from training data. However, limited vocabulary size
increases the amount of unknown words. The OOV problem is that the unknown words make language processing
inaccurate. Subword tokenization refers to dividing text into meaningful tokens. Subword tokenization assumes that a
word consists of multiple pieces of subword. This method makes rare words decomposed into subwords to avoid OOV
problems. It also allows language models to have a reasonable size of vocabulary. Unigram language model tokenization
shows superior performance among a variety of subword tokenization methods. This tokenization makes vocabulary by
using the probability of subword occurrence given by training corpus. Experimental results with Korean corpus of
KsponSpeech show that unigram language model-based subword tokenization leads to significant improvements on n-
gram based language modeling.
RS5 Regular Session: High Performance Computing Systems
Tuesday, 22nd December 2020, 10:45 – 12:00
Chair: Yong Hwan Lee (Wonkwang University, Korea)
RS5-1 Bank and buffer system design for low power L1 cache memory EEECS217
Bosung Jung Gyeongsang National University
Junghoon Lee Gyeongsang National University
Today, with the advent of the 4th industrial revolution, IoT (Internet of Things) systems are rapidly developing.
Accordingly, with the advent of various large-capacity applications, low-power and high-performance memory for
computing systems is being demanded. In this paper, we propose an effective structure to reduce the energy consumption
of the L1 cache memory, which occurs most frequently in the cache system. The proposed cache memory system is largely
composed of two parts, the L1 main cache memory and the buffer cache. The main cache memory is 2-banks, and each
bank consists of a 2-way association map. In the buffer cache operation, data is copied with the access algorithm proposed
in this paper when an access success occurs from the L1 main cache memory. According to our simulation, the proposed
L1 cache system can reduce the energy-delay product by around 55%, compared with 4-way set associative cache memory.
EEECS2020 Program & Abstracts
21
RS5-2 PCM Buffer Cache on PCM Hybrid Memory for IoT Devices EEECS210
Su-Kyung Yoon Jeonbuk National University
In this paper, we present a technique for using PCM as a buffer cache in DRAM/PCM hybrid memory for IoT devices.
In this system, a small amount of PCM is introduced as a write buffer cache to reduce system stability and power
consumption due to DRAM buffer usage. The proposed technique records data attributes for the data requested to be
written, and a write block is created by analyzing the association of data and grouping them. As results of experiments,
the proposed system can reduce the energy consumption by 7.5%. compared to conventional system.
RS5-3 Data Management Techniques for Graph Processing EEECS211
Su-Kyung Yoon Jeonbuk National University
Recently, the use of memory-intensive applications such as graph processing is increasing. Memory-intensive
applications require a large amount of main memory for processing. However, DRAM, which is used as a basic
component of main memory, has reached its scaling limit. In this paper, to solve this problem, we design a method of
prefetching data by predicting data to be requested in advance in PCM-based hybrid memory. According to the
experimental results, the proposed system reduces the execution time by 4.9%.
RS5-4
Implementation of KD-tree Builder for Real-time Ray and Sound Tracing
Applications
EEECS225
Woonam Chung Sejong University
Sukwon Choi Sejong University
Woochan Park Sejong University
Cheong-Ghil Kim Namseoul University
In order to support the dynamic scene rendering based on ray and sound tracing algorithms, an optimized data
structure called acceleration structure (AS), such as KD-tree structure, should be built for every frame at run-time. In this
paper, for the efficient implementation of the AS builder hardware, a novel KD-tree build algorithm with computational
complexity of Nlog(N) has been divided into software part and hardware part. In the software part, sorted information of
triangles is generated. In the hardware part, the KD-tree for each image frame is built with the sorted triangle information.
According to experimental results, the proposed hardware architecture could reduce the tree build latency by about
50%~150% compared to the software-only approach.
RS5-5 A Design of Inverse Cosine Calculator for Sound Tracing Applications EEECS226
JinYoung Lee Sejong University
Woochan Park Sejong University
Cheong-Ghil Kim Namseoul University
Sound tracing is a technology that generates sound by finding a sound propagation path from a sound source to a
listener and calculating it in real-time. These calculations have to deal with complex operations such as trigonometric
functions. In this paper, the inverse cosine function with a very small error value is implemented in hardware by
effectively constructing the table method. Since the sound tracing application field does not require relatively high
accuracy, a method that can obtain an approximation value within a reasonable allowable range with a small hardware
size was proposed.
Program & Abstracts EEECS2020
22
RS6 Regular Session: Intelligent Signal and Image Processing 2
Tuesday, 22nd December 2020, 14:45 – 16:00
Chair: Kyoungro Yoon (Konkuk University, Korea)
RS6-1 Evaluation on Single Image Reflection Removal EEECS219
Hyochang Ahn Far East University
Yong-Hwan Lee Wonkwang University
An undesired negative image has occurred in photographs taken across partial reflections such as glass windows and
electronic display. Efficient removing reflection given a single image is emerging in recent research on image processing.
This paper discusses and evaluates two published image reflection removal methods, and compares the performance of
the processing time and quality of those schemes with common datasets. As benchmarking tests are presented, and we
propose to modify one of the methods to reduce the run-time with small effects on similar image quality.
RS6-2
Deep Learning Based Viewport Super Resolution by Using 2d Warping
Method for 360-Degree Video
EEECS249
Seongbae Rhee Kyung Hee University
Seonghwan Park Kyung Hee University
Kyuheon Kim Kyung Hee University
Recently, 360-degree video has been increasingly used in monitoring systems. Since a 360-degree video requires a
considerably high transmission bandwidth, it has a limitation in providing a network based monitoring services. In order
to overcome this limitation, this paper proposes the method of transmitting a relatively lower-quality video and
reproducing it as a high-resolution video through deep learning Super Resolution(SR) network. Although many SR
networks have shown significant performance in a legacy 2D video, the application of SR to 360-degree video is limited
due to the radial distortion of the wide-angle lens. Therefore, in this paper, we propose the spatial transformation suitable
for SR network to overcome radial distortion.
RS6-3 Subjective quality evaluation of point clouds compared to natural video EEECS234
Aram Baek Hanbat National University
Haechul Choi Hanbat National University
Point clouds are continually evolving as a representative way of expressing 3D contents. In order to evaluate the
quality of these point clouds, various objective or subjective quality evaluation methods are being studied. In this paper,
a subjective quality evaluation was performed in comparison with natural video to compare the several quality of the
point clouds. As the experimental results, the measurement for each quality of the point clouds is compared and analyzed.
RS6-4
Frequency-Tuned Spectrogram with Residual Convolutional Neural
Networks for Acoustic Scene Classification in Multi-Device Environment
EEECS242
Soonshin Seo Sogang University
Changmin Kim Sogang University
Donghyun Lee Sogang University
Hosung Park Sogang University
Hyunsoon Son Sogang University
Gyujin Kim Sogang University
EEECS2020 Program & Abstracts
23
Ji-Hwan Kim Sogang University
In this paper, we proposed an acoustic scene classification system in a multi-device environment using frequency-
tuned spectrogram and residual convolutional neural networks. It aims to maximize the distance of training samples
between acoustic scene classes and to minimize the distance of training samples between multi-devices. The inter-class
and inter-device standard deviations on the frequency axis are calculated by using the training dataset. These statistics are
used to tune the log-mel spectrogram. Frequency-tuned spectrogram and the log-mel spectrogram are used as input
features of two-pathway residual convolutional neural networks. Our proposed system achieved an overall accuracy of
71.7% for the TAU Urban Acoustic Scenes 2020 Mobile dataset.
RS6-5 Comparison of CNN and YOLO for Object Detection EEECS220
Yong-Hwan Lee Wonkwang University
Youngseop Kim Dankook University
Object detection has to play an important role in the field of comupter vision. Many kind of research have rapidly
increased along with applying neural network and its improved structures since 2012. There are representative object
detection algorithms, which are convolutional neural networks and YOLO. This paper presents two representative
methods, based on CNN and YOLO which solve the problem of CNN bounding box. We have compared the performance
of two methods in terms of accuracy, speed and cost. Compared with the latest advanced solution, YOLO v3 achieves a
good trade-off between speed and accuracy.
RS7 Regular Session: Immersive Display, Network & Contents
Tuesday, 22nd December 2020, 13:15 – 14:30
Chair: Jin Young Kim (Kwangwoon University, Korea)
RS7-1 ANN-based Body Temperature Monitoring for Home Healthcare System EEECS252
Kyu-Nam Choi Gachon University
Taeg-Keun Hwangbo Gachon University
One of the most important indicators in home healthcare systems is body temperature. Many diseases are the first to
cause changes in human body temperature. Therefore, continuous observation of body temperature is the starting point
of home healthcare in the fastest way to detect disease. Conventional methods of temperature measurement are
inconvenient for the elderly and for those living alone. To solve the above problem, a temperature measurement system
using a smart mirror system equipped with a camera, a contactless temperature camera and a laser distance sensor is
proposed. By detecting the face based on deep learning, automatically tracking the target of measurement, and using the
temperature camera through the ANN-based adaptive distance compensation, accurate temperature measurement is
possible in real time. The results of the proposed system show very good performance. The experiment was carried out
for a week in a general home, and as a result of comparing the axillary body temperature measured with a mercury
thermometer as a reference standard, the accuracy was 92%, showing high diagnostic accuracy.
RS7-2 Promoting of Online PBL For Collaborative Skills Enhancement EEECS250
Santichai Wicha Mae Fah Luang University
Panarumol Temdee Mae Fah Luang University
This paper investigates a Problem based learning (PBL) model which focuses on how to improve the collaborative
skills construction in online classroom. The empirical study conducted with Business intelligence class from School of
Information technology, Mae Fah Luang University, Thailand. The results of this study indicates that online PBL as a
Program & Abstracts EEECS2020
24
learning environment that supported self-directed and collaborative learning skills during they worked on online PBL
program projects.
RS7-3 CIA And IoT Communication Models EEECS251
Chayapol Kamyod Mae Fah Luang University
Agriculture is a major contributor to the Thailand economy. Accordingly, young, and smart farmers in Thailand have
gained more interest and applied various IoT solutions and applications to enhance their production’s quality and quantity.
Nevertheless, general IoT solutions that are available in the market cannot provide full functionalities to serve different
types of agricultural processes. Therefore, various IoT solutions from different vendors were applied. Some farmers used
customized IoT solutions for their specific production processes such as fertilization and irrigation. Accordingly, various
IoT deployments can create complexity and vulnerability. Therefore, security of the system should also be considered at
this beginning employment state to reduce possible risks when scaling the production processes or system. This paper
examines security of IoT communication models by using the commonly used framework called the Cybersecurity Cube
which represents three principles of information security or CIA triad. The study represents various vulnerabilities of
different IoT communication models.
RS7-4
A Study of Proxy Based Information Crawling on Multi-Screen Digital
Signage
EEECS237
Heechan Yang Namseoul University
Daeseung Park Namseoul University
Sangwoon Lee Namseoul University
Cheong Ghil Kim Namseoul University
Chayapol Kamyod Mae Fah Luang University
With the advancement of network and multimedia technologies, digital signage via the Internet or Cloud has been a
dramatic transformation. As one of emerging commercial advertisement channel, it is important to have the functions of
collecting the data generated by sensors or other edge nodes and crawling the user request for a corresponding documents
in the distributed systems. In this paper, we survey the structure of the recent network-based digital signage platform and
introduce a new proxy based information crawing system architecture for multi-screen digital signage.
RS8 Regular Session: Artificial Intelligence for Embedded Systems
Monday, 21st December 2020, 15:45 – 17:00
Chair: Youngmo Kim (Soongsil University, Korea)
RS8-1
Implementation of Contactless Vehicular Ambient Lighting Control System
Based on the Machine Learning Method
EEECS206
Sangyub Lee Korea Electronics Technology Institute
Youngchan Kim Korea Electronics Technology Institute
This paper is about the implementation of a system that controls ambient lighting, which is an emotional lighting in
a vehicle. Ambient lighting is an in-vehicle lighting assistant that provides a function to change the color to proper the
user's emotions in car. Existing ambient lighting control allows the user to access the central control display or head unit
system and activate the corresponding function within the system. In this paper, we propose a method of implementing
an in-vehicle emotional lighting control system based on a machine learning algorithm that can easily recognize the user's
intention using a simple gesture without the user's direct device manipulation using a magnetic field sensor module.
EEECS2020 Program & Abstracts
25
RS8-2
Development of KNN-based Haar face detection system using low-power
neuromorphic chip
EEECS212
YoungChan Kim Korea Electronics Technology Institute
YeoWool Lee Korea Electronics Technology Institute
Inpyo Cho Korea Electronics Technology Institute
SangYub Lee Korea Electronics Technology Institute
In this paper, we propose a facial recognition system that can be used in a low-power environment such as an
embedded system by using a neuromorphic chip equipped with a KNN matching function. The Haar cascade classifier
method proposed by Paul Viola and Michael Jonse extracting the Haar features of an image to detect face has a lot of
real-time computation, so it is difficult to find a face in a video feed with a high frame rate in a low-clock MCU. Therefore,
in this study, we introduce a system that improves traditional Haar cascade classifier method by vectorizing the human
facial Haar features on a neuromorphic chip connected to an embedded environment, and calculating them at once. In
addition, this system can be applied to fields requiring face recognition (for example, driver's front gaze system) and
shows the possibility of realization through prototype examples.
RS8-3
Uninserted Error Checking Technique Using RBF Tester for Low Power Low
Cost PCB Tester
EEECS213
InPyo Cho Korea Electronics Technology Institute
YeoWool Lee Korea Electronics Technology Institute
JaeKyu Lee Korea Electronics Technology Institute
SangYub Lee Korea Electronics Technology Institute
The PCB inspector is a device that checks whether the PCB is correctly manufactured at the end of the SMT line. In
this paper, we propose an uninterpolated error inspection technique that can be used for low-power, low-cost PCB
inspection machines. In high-end equipment, high-performance lighting, cameras, and processors can be used to perform
PCB inspections, but in low-end equipment, sufficient lighting, cameras, and processors cannot be used, so an appropriate
algorithm and accelerator must be used. Using the algorithm and RBF accelerator, an image of 10,000 pixels is inspected,
and even Arduino class hardware can perform the inspection in 30 milliseconds on average.
RS8-4
A Study on the Structure of Learning Image Dataset for Object Recognition
in Construction Sites based on Transfer Learning
EEECS214
Jaekyu Lee Korea Electronics Technology Institute
Chulgoo Kim Korea Electronics Technology Institute
InPyo Cho Korea Electronics Technology Institute
Sangyub Lee Korea Electronics Technology Institute
In this paper, we conducted a study on the structure of the learning image data set for object recognition in
construction sites based on transfer learning. In particular, we explored the structure of image data sets such as VOC, MS
COCO and KITTI, and we conducted a study on the composition of the data file for describing image information. The
construction site images that we collected were converted to the KITTI data format, and we created a model to recognize
objects in the construction site using the transformed KITTI image data set format based on transfer learning.
RS8-5
Development of a digital twin sensor platform for disaster monitoring and
management
EEECS221
Changsuk Yoon Korea Electronics Technology Institute
Minsang Yu Korea Electronics Technology Institute
Program & Abstracts EEECS2020
26
Young-Han Kim Korea Electronics Technology Institute
Hyun-Seok Ahn Korea Electronics Technology Institute
Tae-Beom Lim Korea Electronics Technology Institute
Recently, as IoT systems have been widely adopted in various industrial fields, continuous and immediate data
collection in the field has become possible. This enables more accurate analysis and prediction that reflects reality
compared to existing theoretical simulations. Now, it is possible to effectively monitor and predict the various disasters
that occur continuously and frequently. The disaster monitoring and management system is expected to have great
effectiveness in economic and social aspects. However, sufficient research has not been conducted on the digital twin
system structure and management software framework for effective application. In this paper, the sensor processing
platform constructing the digital twin infrastructure for effective disaster monitoring and management is presented.
SS1 Special Session: Smart Multimedia Applications
Monday, 21st December 2020, 15:45 – 17:00
Chair: Jeong-geun Kim (Yonsei University, Korea)
SS1-1
A Contactless Personal Health Monitoring System Using Raspberry Pi Based
Smart Mirror
EEECS236
Jiwoong Han Namseoul University
Changhoon Kim Namseoul University
Daeseung Park Namseoul University
Sangwoon Lee Namseoul University
Cheong Ghil Kim Namseoul University
The internet of things (IoT) is rapidly expanding its application range with the commercialization of 5G service. In
this paper, we propose a Raspberry Pi based samrt mirror for a contactless personal health monitoring, which may make
our lives more convenient by adding IoT functions to our daily used item of mirrors. The smart mirror to be implemented
will use a technology of measuring bio-signal in a contactless manner using Lepton thermal imaging sensor, remote
photoplethysmography (rPPG) with regular color video camera, and so on. The general features of the smart mirror will
be provided on the basis of open source MagicMirror including news, weather, calendar, and multimedia service.
SS1-2
Problem-Based Learning Model for Students Knowledge Acquisition of
Logical Design for Interactive Programming Course
EEECS240
Waralak Chongdarakul Mae Fah Luang University
Khwunta Kirimasthong Mae Fah Luang University
Developing the learning approach supportive to the students in the field of digital technology implementation for
business study has been challenged. This research employed Problem Oriented Project Based Learning (POPBL) principle
to construct the multidisciplinary learning model to promote knowledge acquisition of the first-year undergraduate
students in DTBI curriculum. The student performance evaluation is designed for project works and problem-solving
assessments of which students can eventually accomplish outcome of the real business initiation. The assessment result
presents the evaluation score and useful feedback for improvement.
SS1-3
An Assistive Technology for the Elderly Farmers: A Case Study of a
Cymbidium Orchid Caring System
EEECS245
Kesmanee Runjarean Mae Fah Luang University
EEECS2020 Program & Abstracts
27
Nattanicha Thammakhankhom Mae Fah Luang University
Warisara Luetan Mae Fah Luang University
Prasara Jakkaew Mae Fah Luang Universityty
Sirichai Hemrungrote Mae Fah Luang University
Most elderly orchid farmers in Thailand currently rely on a traditional farming process, which requires intensive
care with a lot of time spending. To improve efficiency, many pieces of the literature suggest using wireless sensor
networks as an assistive technology in controlling the farming's environment, such as temperature, humidity, soil moisture.
However, the deployment of such a system with its initially high investment has not gained much popularity among those
elderly farmers. Unlike other previous works in this paper, we mainly focus on designing a smart care system with an
easy-to-use user interface design. We aim to investigate the significance of a user-friendly application and the efficiency
improvement obtained from adopting wireless sensor technology over a traditional approach.
SS1-4
The Combined Model of Blended Learning with Flipped Classroom: A New
Learning Strategy for Covid-19 Pandemic Situation
EEECS246
Worasak Rueangsirarak Mae Fah Luang University
Punnarumol Temdee Mae Fah Luang University
In Thailand, blended learning was introduced to be the strategy for teaching management during the COVID-19
pandemic situation. Therefore, the proposed combined model of blended learning with flipped classroom was designed
to help the lecturers to organize the online teaching for their students. The idea is to focus on two event activities, one is
the self-learning through the LMS and another is the wrapped up lecture by the facilitator. Additionally, the class is
conducted with both synchronous and asynchronous mode. The case study is conducted with the Information Technology
students who enrolled in the Cloud Computing course. As the results, the proposed teaching model can enhance the
students’ knowledge retention more than those of the traditional classroom with the t-Test analysis show its significance
of P-value at 0.014<0.05.
SS1-5
Applying Problem-Based Learning with Direct Instruction Improve Students
Critical Thinking and Teamwork in Thai Undergraduates
EEECS247
Nilubon Kurubanjerdjit Mae Fah Luang University
Soontarin Nupap Mae Fah Luang University
This study aimed to assess the effectiveness of 1) critical thinking and 2) collaborative learning skills of Thai
undergraduate students in a sample of Mae Fah Luang University in Chiang Rai, Thailand. The case study focused on 42
Digital Technology for Business Innovation (DTBI) major in school of Information Technology students who enrolled in
“Digital Business Initiation in Practices” course. We endeavoured to apply problem-based learning strategy in order to
stimulate critical thinking and collaborative learning to work as a team of the students. The results revealed the application
of PBL strategy improved the critical thinking of students significantly as well as the high performance on their works
and their roles playing within their team in a collaborative learning environment.
SS2 Special Session: Immersive Content Display
Monday, 21st December 2020, 15:45 – 17:00
Chair: Sunghwa Lim (Namseoul University, Korea)
SS2-1
DUET-Based Separation of Multiple User Signals for Digital Drone Signage
Systems
EEECS205
Isaac Sim Kwangwoon University
Program & Abstracts EEECS2020
28
Young Ghyu Sun Kwangwoon University
Soo Hyun Kim Kwangwoon University
Donggu Lee Kwangwoon University
Jiyoung Lee Kwangwoon University
Jin Young Kim Kwangwoon University
In this letter, we study a scenario based on degenerate unmixing estimation technique (DUET) that separates original
signals from mixture of FHSS signals with two antennas. We have shown that the assumptions for separating mixed
signals in DUET can be applied to drone based digital signage recognition signals and proposed the DUET-based
separation scheme (DBSS) to classify the mixed recognition drone signals by extracting the delay and attenuation
components of the mixture signal through the likelihood function and the short-term Fourier transform (STFT). In
addition, we propose an iterative algorithm for signal separation with the conventional DUET scheme. Numerical results
showed that the proposed algorithm is more separation-efficient compared to baseline schemes. DBSS can separate all
signals within about 0.56 seconds when there are fewer than nine signage signals.
SS2-2 A Study on The Performance Evaluation of Webxr Device Api EEECS223
Daehyeon Lee Kwangwoon University
Munyoung Lee Kwangwoon University
Kye-Dong Jung Kwangwoon University
Soonchul Kwon Kwangwoon University
Recently, with the development of various virtual environment technologies, they are fused with reality or are
defining and researching a new reality. In particular, research on extended reality is being actively conducted in the web
environment. However, the current released tentative version has poor stability. Therefore, in this paper, a study was
conducted on the performance evaluation of WebXR Device API. For the experiment, a real model, the same 3D model,
and a ground environment for the match rate comparison were made, and the match rate between the real model and the
3D model was compared. As a result of the experiment, 92.6%, 88.9%, and 88.8% of the results were confirmed from the
front, left, and right.
SS2-3 A Study on Digital Signage Data Transmission System using RDS EEECS224
SangWoon Lee Namseoul University
In this paper, we propose a transmission system configuration method using FM radio RDS (Radio Data System) as
a method for transmitting data for expression to digital signage. FM RDS transmits data using an FM broadcasting
network that is used worldwide, and has a feature that can reliably transmit data even in natural disasters such as
earthquakes and typhoons. When data is transmitted to a digital signage using this characteristic, it is possible to transmit
important messages including a disaster warning even in a situation in which the data channel over the communication
network transmitted through the digital signage is paralyzed.
In recent years, we have been experiencing an explosive expansion of artificial intelligence (AI) technology in a
number of applications. Furthermore, we do not doubt that the impact of AI technology on our future world will exceed
SS2-4
A Study of Smart Contents Management System Using Machine Learning
for Digital Signage
EEECS235
Daeseung Park Namseoul University
Jiwoong Han Namseoul University
Sangwoon Lee Namseoul University
Cheong Ghil Kim Namseoul University
EEECS2020 Program & Abstracts
29
our imagination. In this work, we introduce a system architecture of data driven smart digital signage using machine
learning. Variety of datasets may come not only in real time from sensors around the digital display but also from the
contensts management system. The proposed the digital signage contents management system using machine learning is
expected to allow different content to be scheduled auttonomously and displayed at different times according to the
situation. As a result, this improvements eventually make smart decisions to make better services for customers.
In this study, the concept of Cross Reality (eXtended Reality) was used to build a virtual production system that
could be applied to superior imaging and to produce special effects images used in actual TV dramas. In the process, the
advantages of real-time rendering were actively revealed, enabling directors and actors to check the video results in real-
time, and ensuring that camera matching in the second half of the work can be carried out smoothly. This study is
meaningful as a success story to confirm that introducing a virtual production system helps to maintain the highest quality
of the video while meeting the production schedule.
SS2-5
System Modeling and Implementation of XR Environment -Focused on
VFX Production for TV Drama-
EEECS222
Jungwoon Park Gachon University
Seokhee Oh Gachon University
Program & Abstracts EEECS2020
30
7 Author Index
A
- Ahn, Hyochang: RS3-2, RS6-1
- Ahn, Hyun-Seok: RS8-5
- An, Byoungman: RS3-1
B
- Baek, Aram: RS6-3
C
- Cho, InPyo: RS8-2, RS8-3, RS8-4
- Choi, Haechul: RS3-3, RS6-3
- Choi, Kyu-Nam: RS7-1
- Choi, Sukwon: RS5-4
- Chongdarakul, Waralak: SS1-2
- Chung, HaeKyung: RS2-2, RS2-4
- Chung, Woonam: RS5-4
H
- Han, Jiwoong: SS1-1, SS2-4
- Hemrungrote, Sirichai: SS1-3
- Hwangbo, Taeg-Keun: RS7-1
J
- Jakkaew, Prasara: SS1-3
- Jang, Seyoung: RS1-4
- Ji, Seunghyun: RS1-5
- Joe, YongJoon: RS1-1
- Jung, Bosung: RS5-1
K
- Kamyod, Chayapol: RS7-3, RS7-4
- Kim, Changhoon: SS1-1
- Kim, Changmin: RS6-4
- Kim, Cheong-Ghil:
RS5-4, RS5-5, RS7-4, SS1-1, SS2-4
- Kim, Chulgoo: RS8-4
- Kim, Gyujin: RS3-5, RS6-4
- Kim, Ji-Hwan: RS3-5, RS6-4
- Kim, Jin Young: SS2-1
- Kim, Kyuheon: RS3-4, RS6-2
- Kim, Seok-Yoon:
RS1-2, RS1-3, RS1-4, RS1-5
- Kim, Seong-Bo: RS1-1
- Kim, Soo Hyun: SS2-1
- Kim, YoungChan: RS8-1, RS8-2
- Kim, Young-Han: RS8-5
- Kim, Youngmo: RS1-2, RS1-3, RS1-5
- Kim, Youngseop: RS3-1, RS6-5
- Kirimasthong, Khwunta: SS1-2
- Ko, JangHyok: RS 2-2, RS2-4
- Kurubanjerdjit, Nilubon: SS1-5
- Kwon, Daehyeok: RS3-3
- Kwon, Hyukmin: RS3-4
- Kwon, Soonchul: SS2-2
L
- Lee, Daehyeon: SS2-2
- Lee, Donggu: SS2-1
- Lee, Donghyun: RS6-4
- Lee, Jaecheng: RS1-2
- Lee, Jaekyu: RS8-3, RS8-4
- Lee, JinYoung: RS5-5
- Lee, Jiyoung: SS2-1
- Lee, Junghoon: RS5-1
- Lee, SangWoon:
RS7-4, SS2-3, SS1-1, SS2-4
- Lee, SangYub:
RS8-1, RS8-2, RS8-3, RS8-4
- Lee, Se Won: RS2-1, RS2-5
- Lee, Seungmin: RS2-3
- Lee, YeoWool: RS8-2, RS8-3
- Lee, Yong-Hwan: RS3-2, RS6-1, RS6-5
- Lim, Sung-Hwa: RS2-1, RS2-5
- Lim, Tae-Beom: RS8-5
- Luetan, Warisara: SS1-3
N
- Nupap, Soontarin: SS1-5
O
- Oh, Seokhee: SS2-5
P
EEECS2020 Program & Abstracts
31
- Park, Byeongchan: RS1-2, RS1-3
- Park, Daeseung: RS7-4, SS1-1, SS2-4
- Park, Hosung: RS3-5, RS6-4
- Park, Jungwoon: SS2-5
- Park, Seonghwan: RS3-4, RS6-2
- Park, Woochan: RS5-4, RS5-5
R
- Rhee, Seongbae: RS3-4, RS6-2
- Rueangsirarak, Worasak: SS1-4
- Ruizive, Ulugbek: RS1-3
- Runjarean, Kesmanee: SS1-3
S
- Seo, Jeongil: RS3-4
- Seo, Soonshin: RS6-4
- Shin, Dong-Myong: RS1-1
- Sim, Isaac: SS2-1
- Son, Hyunsoon: RS3-5, RS6-4
- Sun, Young Ghyu: SS2-1
T
- Temdee, Punnarumol: SS1-4
- Thammakhankhom, Nattanicha: SS1-3
- Temdee, Panarumol: RS7-2
W
- Wicha, Santichai: RS7-2
Y
- Yang, Heechan: RS7-4
- Yoo, Injae: RS1-2
- Yoon, Changsuk: RS8-5
- Yoon, Su-Kyung: RS5-2, RS5-3
- Yu, Minsang: RS8-5