SEULKI LEE (이슬기)

Announcement

Our research group, EAI Lab (Embedded Artificial Intelligence Lab), is looking for new members (MS/PhD/Intern/PostDoc) who are interested in systems and AI research in 2024/2025. Please visit our lab homepage for more information.

About

Seulki Lee is an assistant professor at the Department of Computer Science and Engineering (CSE) and Artificial Intelligence Graduate School (AIGS) at UNIST (Ulsan National Institute of Science and Technology), where he leads EAI Lab (Embedded Artificial Intelligence Lab). He earned his Ph.D. in Computer Science at the Department of Computer Science, the University of North Carolina at Chapel Hill (UNC), advised by Dr. Shahriar Nirjon. His research aims to make resource-constrained real-time and embedded sensing systems capable of learning, adapting, and evolving, which facilitates Embedded Artificial Intelligence (Embedded AI). Prior to starting his Ph.D., he designed sensor system architecture and intelligent sensing algorithms for smartphones at Samsung Electronics. He spent two summers as a research intern at Nokia Bell Labs and Mercedes-Benz Research & Development North America, where he researched machine learning and future transportation systems.

Research Interests

• Embedded Artificial Intelligence (Embedded AI)

  • On-device machine learning on embedded/mobile/IoT devices

  • Embedded Computer Vision, Embedded NLP, Embedded Reinforcement Learning

  • Efficient model (deep learning) inference, training, and adaptation

• Collaborative and Collective Learning Systems

  • Pervasive- and always-learning on connected systems

  • Scalable and autonomous intelligence on networked systems (AIoT, AIoE, and intelligent edge)

  • Mobile computing and sensor data analytics

• Cyber-Physical Systems

  • Control theory

  • Mobile robots and smart sensing systems

Experience

• Assistant Professor: UNIST (Ulsan National Institute of Science and Technology), Ulsan, Korea, 2021 - Present

  • Department of Computer Science and Engineering (CSE)

  • Affiliated with Artificial Intelligence Graduate School (AIGS)

• Research Assistant: University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, 2016 - 2021

  • Department of Computer Science

  • Embedded artificial intelligence, Mobile computing, Machine learning, and Real-time systems

• Research Intern: Nokia Bell Labs, Murray Hill, New Jersey, 2018

  • Mathematics and Algorithm group: Machine learning for large scale face detection/recognition

• Software Engineering Intern: Mercedes-Benz Research & Development North America, Sunnyvale, California, 2017

  • Computer vision (deep learning) and intelligent embedded system for Future Transportation Systems

• Software Developer: Samsung Electronics, Suwon, Korea, 2009 - 2016

  • Sensor R&D Group: SensorHub platform and algorithm for mobile devices (Galaxy S smartphones)

Education

• Ph.D.: Computer Science, University of North Carolina at Chapel Hill, 2021 [hooding graphic]

• M.S.: Computer Science, University of North Carolina at Chapel Hill, 2018

• B.E.: Electrical and Computer Engineering, University of Seoul, Korea, 2009

Publication

2024

• [KDD 2024] “CAFO: Feature-Centric Explanation on Time Series Classification”

  • In the 30th SIGKDD Conference on Knowledge Discovery and Data Mining, Aug 2024, Barcelona, Spain, Aug 2024. To appear.

1. • Jaeho Kim, Seok-Ju Hahn, Yoontae Hwang, Junghye Lee, and Seulki Lee

• [AISTATS 2024] “On-Demand Federated Learning for Arbitrary Target Class Distributions”

  • In the 27th International Conference on Artificial Intelligence and Statistics, Valencia, Spain, May 2024. [pdf][code][video][poster]

1. • Isu Jeong and Seulki Lee

2023

• [NeurIPS 2023] “Softmax Output Approximation for Activation Memory-Efficient Training of Attention-based Networks”

  • In the Thirty-Seventh Annual Conference on Neural Information Processing Systems, New Orleans, LA, USA, Dec 2023. [pdf][slides][code][talk(video)][poster]

1. • Changhyeon Lee and Seulki Lee

• [SenSys 2023] “On-NAS: On-Device Neural Architecture Search on Memory-Constrained Intelligent Embedded Systems”

  • In the 21th ACM Conference on Embedded Networked Sensor Systems, Istanbul, Turkiye, Nov 2023. [pdf][slides][code][talk(video)]

1. • Bosung Kim and Seulki Lee

• [IPSN 2023] "MicroDeblur: Image Motion Deblurring on Microcontroller-based Vision Systems"

  • In the 22th International Conference on Information Processing in Sensor Networks, San Antonio, Texas, USA, May 2023. [pdf][slides]

  • Seulki Lee

  • Best Paper Runner Up

• [IEEE Internet of Things 2023] "Multitask Deep Learning for Human Activity, Speed, and Body Weight Estimation using Commercial Smart Insoles"

  • IEEE Internet of Things, Apr 2023. [pdf]

  • Jaeho Kim, Hyewon Kang, Jaewan Yang, Haneul Jung, Seulki Lee, and Junghye Lee

2022

• [PerCom 2022] "Weight Separation for Memory-Efficient and Accurate Deep Multitask Learning"

  • In the 20th International Conference on Pervasive Computing and Communications, Pisa, Italy, March 2022. [pdf][slides][talk(video)][code]

1. • Seulki Lee and Shahriar Nirjon

2021

• [IPSN 2021] "Deep Functional Network (DFN): Functional Interpretation of Deep Neural Networks for Intelligent Sensing Systems"

  • In the 20th International Conference on Information Processing in Sensor Networks, Nashville, USA, May 2021. [pdf][slides][talk(video)][code]

1. • Seulki Lee and Shahriar Nirjon

• [Ph.D. Dissertation] “Enabling Deep Intelligence on Embedded Systems”

  • The University of North Carolina at Chapel Hill, May 2021. [pdf]

  • Seulki Lee

2020

• [AIChallengeIoT 2020] (Workshop) "Learning in the Wild: When, How, and What to Learn for On-Device Dataset Adaptation"

  • In the 2nd International Workshop on Challenges in Artificial Intelligence and Machine Learning for Internet of Things, Virtual Event, Japan, November, 2020. [pdf][slides][talk(video)]

1. • Seulki Lee and Shahriar Nirjon

   • Best Paper Award [link]

• [MobiSys 2020] "Fast and Scalable In-memory Deep Multitask Learning via Neural Weight Virtualization"

  • In the 18th ACM International Conference on Mobile Systems, Applications, and Services, Toronto, Canada, June 2020. [pdf][slides][talk(video)][code]

1. • Seulki Lee and Shahriar Nirjon

• [RTAS 2020] "SubFlow: A Dynamic Induced-Subgraph Strategy Toward Real-Time DNN Inference and Training"

  • In the 26th IEEE Real-Time and Embedded Technology and Applications Symposium, Sydney, Australia, April 2020. [pdf][slides][talk(video)][code]

1. • Seulki Lee and Shahriar Nirjon

2019

• [IMWUT 2019/UbiComp 2020] “Intermittent Learning: On-Device Machine Learning on Intermittently Powered System”

  • Proceedings of ACM on Interactive, Mobile, Wearable, and Ubiquitous Computing, Cancun, Mexico, September 2020. [pdf][slides][talk(video)][code]

1. • Seulki Lee, Bashima Islam, Yubo Luo, and Shahriar Nirjon

   • Best Presentation Award [link]

• [SenSys 2019] “Neuro.ZERO: A Zero-Energy Neural Network Accelerator for Embedded Sensing and Inference Systems”

  • In the 17th ACM Conference on Embedded Networked Sensor Systems, New York, USA, November 2019. [pdf][slides][talk(video)][code]

1. • Seulki Lee and Shahriar Nirjon

• [RTAS 2019] (WIP) "Time-Aware Deep Intelligence on Batteryless Systems"

  • In the 25th IEEE Real-Time and Embedded Technol-ogy and Applications Symposium, Montreal, Canada, April 2019. [pdf]

1. • Bashima Islam, Seulki Lee, and Shahriar Nirjon

• [IPSN 2019] (Poster) "On-device Training From Sensor Data on Batteryless Platforms"

  • In the 18th International Conference on Information Processing in Sensor Networks, Montreal, Canada, April 2019. [pdf]

1. • Bahsima Islam, Yubo Luo, Seulki Lee, and Shahriar Nirjon

2018

• [DCOSS 2018] "Knowledge Transfer Between Embedded Controllers"

  • In the 14th International Conference on Distributed Computing in Sensor Systems, New York, USA, June 2018. [pdf][slides][demo(video)]

1. • Seulki Lee and Shahriar Nirjon

Teaching

• Advanced Machine Learning (Instructor): Fall 2024

• Introduction to Algorithms (Instructor): Spring 2024, Spring 2023, Spring 2022

• AI Framework Design and Implementation (Instructor): Fall 2023, Fall 2021

• Principles of Programming Languages (Instructor): Fall 2022

• Data Structures (Teaching Assistant), invited by Dr. David Stotts: Spring 2018

• Foundation of Programming (Recitation Lecturer), invited by Dr. Ketan Mayer-Patel: Fall 2017

• Mobile Computing Systems (Teaching Assistant), invited by Dr. Shahriar Nirjon: Spring 2017

Honors and Awards

• Best Paper Award: AIChallengeIoT 2020 [link]

• Best Paper Runner Up: IPSN 2023

• Best Presentation Judges Award: UbiComp/ISWC 2020 [link]

Services

• TPC: AAAI 2025/2024/2023, IEEE PerCom 2023, ICMU 2023 (The 14th International Conference on Mobile Computing and Ubiquitous Networking), ACM SenSys 2019 (posters and demos)

• Reviewer: KDD 2025/2024, ACM Transactions on Architecture and Code Optimization 2023, ACM Conference on Computer-Supported Cooperative Work and Social Computing 2021 (CSCW 2021), ACM Interactive, Mobile, Wearable and Ubiquitous Technologies 2021, 2022 (IMWUT 2021, 2022), IEEE Transactions on Computers 2021, IEEE Transactions on Emerging Topics in Computing 2020, IEEE Transactions on Mobile Computing 2020, IEEE/ACM CHASE 2020

Invited Talks

• Nov 2023: AICON Gwangju 2023, Korea, "Image Motion Deblurring on Microcontroller-based Vision Systems"

• Aug 2023: Hyundai R&D Forum 2023 (2023 현대자동차 R&D 기술포럼), Korea, "SDV (Software-Defined Vehicle) Technology Trend"

• Jan 2023: ICOIN (The 37th International Conference on Information Networking) 2023, Thailand, "Enabling Deep Intelligence on Embedded Systems"

• Jun 2022: KCC (Korea Computer Congress) 2022, Korea, "Learning in the Wild"

• Feb 2022: KIISE Computer System Society 2022, Korea, "Enabling Deep Intelligence on Embedded Systems"

• Aug 2021: ETRI (Electronics and Telecommunications Research Institute), Korea, "Enabling Deep Intelligence on Embedded Systems"

• Jun 2021: ETRI (Electronics and Telecommunications Research Institute), Korea, "Intermittent Learning"

• Nov 2020: UNIST (Ulsan National Institute of Science and Technology), Korea, "Towards Embedded Artificial Intelligence (EAI)"

• Nov 2020: DGIST (Daegu Gyeongbuk Institute of Science and Technology), Korea, "Towards Embedded Artificial Intelligence (EAI)" and "Introduction to Deep Learning"

• Oct 2020: GIST (Gwangju Institute of Science and Technology), Korea, "Towards Embedded Artificial Intelligence (EAI)" and "Introduction to Deep Learning"

• Aug 2020: Yonsei University, Korea, "Towards Embedded Artificial Intelligence (EAI)"

• Jun 2020: Stony Brook University, "Fast and Scalable In-memory Deep Multitask Learning via Neural Weight Virtualization"