Wonseon Lim

Wonseon Lim

PhD Candidate at Chung-Ang University. Efficient deep learning for on-device AI.

Wonseon Lim

I am a Ph.D. candidate in Computer Science and Engineering at Chung-Ang University, advised by Prof. Jaesung Lee and Prof. Dae-Won Kim.

My research focuses on designing efficient and adaptive deep learning systems for on-device environments, where models must operate under strict constraints on computation, memory, and latency. I am particularly interested in bridging learning algorithms and deployment-aware optimization to enable scalable, deployable AI across diverse hardware platforms.

Education

  • Ph.D. Candidate in Computer Science and Engineering, Chung-Ang University, 2021 – Present
    • Advisor: Prof. Jaesung Lee, Prof. Dae-Won Kim
  • Visiting Student, Mechanical and Industrial Engineering, University of Toronto, 2024
    • Advisor: Prof. Chi-Guhn Lee
  • M.S. in Computer Science and Engineering, Chung-Ang University, 2019 – 2021
    • Advisor: Prof. Jaesung Lee, Prof. Dae-Won Kim
  • B.S. in Applied Statistics, Chung-Ang University, 2013 – 2019

Research Interests

  • On-Device AI — On-Device Training, Online Continual Learning, Token Pruning
  • Model Optimization — Hardware-Aware Neural Architecture Search, Post-Training Quantization

News

  • [2026.06] Won the Grand Prize (1st Place) at the Nota AI On-Device AI Optimization Competition, held at KCC 2026.
  • [2026.02] CPS-Prompt Paper accepted to CVPR 2026.
  • [2024.03] MixER published in IEEE Access.
  • [2024] Awarded MSIT International Research Training Fellowship; visiting student at the University of Toronto.
  • [2023.09] GTSNet published in Engineering Applications of Artificial Intelligence (EAAI).
  • [2023.07] HAR-LuT-NAS published in IEEE Access.
  • [2022.07] Graves’ Orbitopathy paper published in Scientific Reports.
  • [2022.07] Group-Structure NAS published in Electronics Letters.
  • [2022.05] MTS-ConvNet published in IEEE Access.