Gyuseong Lee

I am a Ph.D. student at CVLAB, KAIST, working on computer vision and generative models.

My research focuses on diffusion- and flow-based generative models, as well as data-centric approaches for vision models.

Prior to joining KAIST, I worked as a Computer Vision Research Scientist at LG Electronics (2024-2026). I received my M.S. in Computer Science from Korea University (2022-2024).

Selected Publications

• Gyuseong Lee*, Wooseok Jang*, Jin Hyeon Kim, Jaewoo Jung, Seungryong Kim.
  Domain Generalization using Large Pretrained Models with Mixture-of-Adapters.
  WACV, 2025.

• Jisu Nam, Gyuseong Lee, Sunwoo Kim, Hyeonsu Kim, Hyoungwon Cho, Seyeon Kim, Seungryong Kim.
  Diffusion Model for Dense Matching.
  ICLR (Oral), 2024.

• Susung Hong, Gyuseong Lee, Wooseok Jang, Seungryong Kim.
  Improving Sample Quality of Diffusion Models Using Self-Attention Guidance.
  ICCV, 2023.

• Junyoung Seo*, Gyuseong Lee*, Seokju Cho, Jiyoung Lee, Seungryong Kim.
  MIDMs: Matching Interleaved Diffusion Models for Exemplar-Based Image Translation.
  AAAI, 2023.

• Jiwon Kim*, Kwangrok Ryoo*, Junyoung Seo*, Gyuseong Lee*, Daehwan Kim, Hansang Cho, Seungryong Kim.
  Semi-Supervised Learning of Semantic Correspondence With Pseudo-Labels.
  CVPR, 2022.

The full publication list is available on Google Scholar

Awards & Service

Outstanding Reviewer: ECCV 2024

Reviewer: CVPR, ICCV/ECCV, AAAI, WACV (2023–present), ICLR (2026-present)