A Researcher & Engineer, learning to share with others
I am a Staff Software Engineer at Samsung Research, with a Ph.D. in Electrical Engineering from KAIST. My research has focused on distributed computing across various stages of the deep learning lifecycle, including distributed training, serving, model personalization, and test-time adaptation.
Currently, my primary focus is on on-device AI technology. I am also a contributor to NNTrainer, an open-source software framework that supports on-device training.
[CV] [LinkedIn] [GoogleScholar] [Github]
▪️ WORKING EXPERIENCE
Samsung Research, Staff Software Engineer
2024.01 ~ Current
NNTrainer [github]
AI Domain Specialist (Oct. 2024)
S/W Award, open source (Nov. 2024)
▪️ EDUCATIONS
Ph. D. Candidate in Electrical Engineering, KAIST
2018. 03 - 2023.08
Advisor: Prof. Chan-Hyun Youn
Research Assitant at Network and Computing Lab (NCL)
Research Interest: Test Time Adaptation, Distributed Deep Learning, Model Personalization, Generative Model
Thesis: A Study on Variational Inference-based Deep Learning Algorithms for Multi-Source Data with Dataset Shift
M. S. in Electrical Engineering, KAIST
2016.03 - 2018.02
Advisor: Prof. Chan-Hyun Youn
Research Assistant at Network and Computing Lab (NCL)
Thesis: A Batch Orchestration Algorithm for Straggler Mitigation of Synchronous SGD in Heterogeneous GPU Cluster
B. S. in Electrical Engineering & Computer Science (Double Majors), Handong Global University (HGU)
2012.03 - 2016.02
Research Assistant at Monolithic Microwave Integrated Circuit Laboratory (MMIC), 2014.06 ~ 2015.12
Graduation with Honor, Summa Cum Laude
▪️ PUBLICATIONS
"Controllable Forgetting Mechanism for Few-Shot Class-Incremental Learning", is accepted to ICASSP 2025! [paper]
K. Paramonov, M. Ozay, E. Yang, J. Moon, U. Michieli, "Controllable Forgetting Mechanism for Few-Shot Class-Incremental Learning", [preprint]
Eunju Yang, Gyusang Cho, and Chan-Hyun Youn "FREEDOM: Label & Domain Information & Source-Free Multi-Source Domain Adaptation for Unsupervised Personalization"[preprint]
Eunju Yang and Chan-Hyun Youn "Temporal Data Pooling with Meta-Initialization for Individual Short-Term Load Forecasting",in IEEE Transactions on Smart Grid (IF: 10.275), vol. 14, no. 4, pp. 3246-3258, July 2023, doi: 10.1109/TSG.2022.3225805.
Eunju Yang, Dong-Ki Kang, and Chan-Hyun Youn, "BOA: batch orchestration algorithm for straggler mitigation of distributed DL training in heterogeneous GPU cluster", The Journal of Supercomputing 2020 (IF: 2.600).
Dong-Ki Kang, Eunju Yang, and Chan-Hyun Youn, "Deep learning-based sustainable data center energy cost minimization with temporal MACRO/MICRO scale management", IEEE Access 2018 (IF: 4.098).
Eunju Yang, Seonghyeon Kim, Chan-Hyun Youn "FPGA-Docker-based MEC Server Framework for Mobile Computing Service with Low-latency" Journal of The Korean Institute of Communication Sciences, Information and Communications Magazine, Vol. 35 No. 11 (2018).
Eunyoung Oh, Duc Canh Le, Eunju Yang, Chan-Hyun Youn "vFPGA to Accelerate Edge Computing Application", Journal of The Korean Institute of Communication Sciences, Information and Communications Magazine Vol. 35 No. 11 (2018).
Eunju Yang and Chan-Hyun Youn, "Individual Load Forecasting for Multi-Customers with Distribution-aware Temporal Pooling", proceed in IEEE International Conference on Computer Communications (INFOCOM) 2021, May, Virtual (acceptance rate: 19.9%) 📄 [slide]
Eunju Yang, Changha Lee, Ji-Hwan Kim, Tuan Manh Tao, and Chan-Hyun Youn “StreamDL: Deep Learning Serving Platform for AMI Stream Forecasting” to appear in proceedings of IEEE International Conference on Data Mining (ICDM) Workshop 2020. 📄
Sunghyun Kim, Eunju Yang, and Chan-Hyun Youn "An Accelerated Edge Computing with a Container and Its Orchestration" International Conference on ICT Convergence (ICTC) 2019 (Excellent Paper Award) 📄
Changha Lee, Minsu Jeon, Eunju Yang, Seonghwan Kim, Chan-Hyun Youn “Imbalanced-Free Memory Selection Scheme Based Continual Learning by Using K-means Clustering” the 10th International Conference on Information and Communication Technology Convergence (ICTC’19).
Eunju Yang, Seong-Hwan Kim, Tae-Woo Kim, Minsu Jeon, Sangdon Park, Chan-Hyun Youn "An adaptive batch-orchestration algorithm for the heterogeneous GPU cluster environment in distributed deep learning system" IEEE International Conference on Big Data and Smart Computing (BigComp 2018) 📄
Taewoo Kim, Eunju Yang, Soyoon Bae, and Chan-Hyun Youn "IDLE: Integrated Deep Learning Engine with Adaptive Task Scheduling on Heterogeneous GPUs" IEEE Region 10 Conference (TENCON) 2018.
Eunyoung Oh, Wan-Gyu Han, Eunju Yang, Jae-Ho Jeong, L. Lemi, Chan-Hyun Youn "Energy-Efficient Task Partitioning for CNN-based Object Detection in Heterogeneous Computing Environment" International Conference on ICT Convergence (ICTC) 2018.
Minsu Jeon, Eunju Yang, Eunyoung Oh, Jeongkyu Park, and Chan-Hyun Youn "A Deterministic Feedback Model for Safe Driving based on Nonlinear Analysis Scheme", EUSPN workshop, Connected & Intelligent Mobility (CIM), 2017 📄
Heejae Kim, Jiyong Han, Seong-Hwan Kim, Jisoo Choi, Dongsik Yoon, Minsu Jeon, Eunjoo Yang, Nhat Pham, Sungpil Woo, Jeongkyu Park, Daeyoung Kim, Chan-Hyun Youn, "IsV2C: an integrated road traffic-network-cloud simulator for V2C connected car services" IEEE International Conference on Services Computing (SCC) 2017 📄
▪️ HONORS & AWARDS
03.2019 - 02. 2022: Global Ph. D. Fellowship Program from National Research Foundation of Korea
05.2019 - 12.2019: Dong-Hwa AI Scholarship from Dong-Hwa Industry Scholarship Foundation
03.2016 - 02. 2022: Government Scholarship for M.S./Ph.D. Students from Korea Government
03.2014 - 02. 2016: National Science & Technology Scholarship from Korea Student Aid Foundation
03.2012 - 02. 2016: Merit-Based Scholarship (Top 1%) from Handong Global University (HGU)
03.2012 - 02. 2016: Undergraduate Full Scholarship from Samsung Scholarship Foundation
10. 2019: Excellent Paper Award from International Conference on ICT Convergence, 2019.
12. 2020: Best Paper Honorable Mention from Coursework Workshop, Advanced Deep Learning Course (AI602, KAIST)
▪️ RESEARCH PROJECTS
07.2022 - present: AI Model Development for Egg Data Processing
03.2021 - present: Core Technology Development for Intelligent Searching and Utilizing Big Data based on DataMap funded by IITP, Korea
10.2019 - 02. 2021: An eXplainable AI (XAI)-based smart energy platform development funded by KEPCO
03.2018 - 12.2019: Developing a Kubernetes-based MEC framework funded by IITP, Korea
06.2018 - 11.2018: Study on Unsupervised Learning & Self Adaptive Model funded by ETRI, Korea
03.2017 - 02.2018: Demand forecasting algorithm development and factor analysis funded by Hyundai Motor Company
06.2016 - 12.2017: A distributed deep learning framework development funded by IITP, Korea
03.2016 - 02.2017: Vehicle-to-Cloud simulator development funded by Hyundai Motor Company
▪️ TEACHING EXPERIENCES & REVIEWER SERVICES
Fall 2019, Teaching Assistant, Service-Oriented Computing (EE614)
Spring 2019, Teaching Assistant, Distributed Computing System (EE613)
Fall 2018, Teaching Assistant, Introduction to Optimization (EE424)
Spring 2018, Teaching Assistant, GPU Programming and Application (EE817)
Fall, 2015, Teaching Assistant, Electronic Circuit 2 (ECE30052)
Spring, 2015, Teaching Assistant, Signal and System (ECE30081)
Spring, 2015, Teaching Assistant, Basic Circuit Lab (ECE20060)
IEEE Transactions on Vehicular Technology
IEEE Transactions on Smart Grid
IEEE Transactions on Network Science and Engineering
IEEE Transactions on Industrial Informatics
The Journal of Supercomputing
IEEE International Conference on Data Mining Workshop