ECE / AI, Seoul Nat'l Univ

Intelligent Comput. imag. Lab.

The Intelligent Computational Imaging Lab (ICL) builds grounded AI that sees, reasons, and acts. Our research follows the full arc of visual intelligence: from physics-aware image formation, through diffusion models for generation, editing, and inverse problems, to vision-language(-action) models that perceive the open world and act within it. We make these models efficient and trustworthy enough to deploy, uncovering subtle anomalies and defects in real-world production and operations. In medicine and science, our work spans scales from radiological imaging and gigapixel pathology to protein and genomic modeling. We publish at leading machine learning and computer vision conferences and in top imaging and signal processing journals. ICL is at the Department of Electrical and Computer Engineering, Seoul National University and is led by Se Young Chun. ICL is also affiliated with the Interdisciplinary Program in Artificial Intelligence, Seoul National University.

ICL is looking for graduate students and undergraduate interns interested in machine learning, signal processing, and computer vision — with applications in generative AI and diffusion models, vision-language and multimodal models, medical and computational imaging, and/or robust image restoration. For more information, click here.

Recent News / Publications (~1 years)

(2026 Sep) Three papers accepted at NeurIPS'26 (1 selected as Spotlight).
- Hoigi Seo* (MS/PhD student), Byung Hyun Lee* (MS/PhD student), Minjun Kim*, Dohyun Mah (MS/PhD student), Jongho Lee, Se Young Chun, "Strong Helps Weak: Directional Cross-Modal Alignment Transfer in Multi-modal LLMs," Accepted at NeurIPS, 2026 (*co-first authors) (Spotlight)
- Junseo Bang (MS/PhD student), Daewon Choi (MS student), Dong Ju Mun (MS/PhD student), Se Young Chun, "FLINT: Coupling Proximal Initialization and Bounded Stochasticity for Flow-Matching Inverse Problems," Accepted at NeurIPS 2026
- Jinho Park, Se Young Chun, Mingoo Seok, "DictLLM: Post-training Compression of Large Language Model with Dictionary Kernels," Accepted at NeurIPS, 2026
(2026 Sep) Se Young Chun gave an invited keynote at ECCV'26 Beyond Euclidean Workshop.
(2026 Aug) Wongi Jeong won an Youlchon AI Star. Congratulations!
(2026 Jun) Justin Yu (CSAI/EE@Stanford), Youngwook Kim (EES/CSE@SNU), Yoonchae Choi (ECE@SNU) will join summer internships at our Lab.
(2026 Jun) One paper accepted at ECCV'26.
- - Ji Ha Jang* (MS/PhD student), Hayeon Kim* (MS/PhD student), Chulwon Lee (MS student), Junghun J. Kim (MS student), Se Young Chun, "HyFL-CLIP: Hyperbolic Fine-Tuning of CLIP for Robust Long-Context Understanding," Accepted at ECCV 2026 (*co-first authors) [report][project]
(2026 Jun) One collaboration paper accepted at MICCAI'26.
- - H S Lee, E S Kim, D M Bappy, H H Kim, Y S Lee, Se Young Chun, J-H Choi, S H Lee, S J Ahn, “Patch-to-Global: Random Patch Diffusion for Globally Consistent Megapixel Artifact Inpainting in Whole Slide Images,” Accepted at MICCAI 2026
(2026 Jun) Our paper, Erasing Thousands of Concepts," has been honored with the CVPR Compute Transparency Champion award - the highest recognition in CVPR 2026's Compute Reporting Initiative. Congratulations! This work was also highlighted at the Popular Science. See this article.
- - Hoigi Seo* (MS/PhD student), Byung Hyun Lee* (MS/PhD student), Jaehyun Cho (MS/PhD student), Sungjin Lim (MS student), Se Young Chun, "Erasing Thousands of Concepts: Towards Scalable and Practical Concept Erasure for Text-to-Image Diffusion Models," CVPR 2026 (*co-first authors)
(2026 May) One paper accepted at ICML'26.
- - Junseo Bang* (MS/PhD student), Dong Ju Mun* (MS/PhD student), Hoigi Seo (MS/PhD student), Seongmin Hong (Postdoc), Se Young Chun, "Triadic Dynamics Aware Diffusion Posterior Sampling for Inverse Problems: Optimizing Guidance and Stochasticity Schedules," Accepted at ICML 2026 (*co-first authors) [report][project]
(2026 Apr) Se Young Chun will serve as an Area Chair of NeurIPS'26.
(2026 Apr) Gwanghyun Kim has joined NVIDIA, USA, as a Research Scientist. Congratulations!
(2026 Apr) Seven papers accepted at CVPR'26 (3 selected as Highlights!).
- - Gwanghyun Kim* (PhD student), Junghun J. Kim* (MS student), Suh Yoon Jeon* (MS/PhD student), Jason Park (MS/PhD student), Se Young Chun, "Human Interaction-Aware 3D Reconstruction from a Single Image," Accepted at CVPR, 2026 (*co-first authors) (Highlight)
  - Hayeon Kim* (MS/PhD student), Ji Ha Jang* (MS/PhD student), Junghun J. Kim (MS student), Se Young Chun, "Uncertainty-guided Compositional Alignment with Part-to-Whole Semantic Representativeness in Hyperbolic Vision-Language Models," Accepted at CVPR, 2026 (*co-first authors) (Highlight)
  - Hoigi Seo* (MS/PhD student), Byung Hyun Lee* (MS/PhD student), Jaehyun Cho (MS/PhD student), Sungjin Lim (MS student), Se Young Chun, "Erasing Thousands of Concepts: Towards Scalable and Practical Concept Erasure for Text-to-Image Diffusion Models," Accepted at CVPR, 2026 (*co-first authors)
  - Wongi Jeong* (MS/PhD student), Hoigi Seo* (MS/PhD student), Se Young Chun, "Training-free, Perceptually Consistent Low-Resolution Previews with High-Resolution Image for Efficient Workflows of Diffusion Models," Accepted at CVPR, 2026 (*co-first authors)
  - Wongi Jeong* (MS/PhD student), Kyungryeol Lee* (MS/PhD student), Hoigi Seo (MS/PhD student), Se Young Chun, "Training-free Mixed-Resolution Latent Upsampling for Spatially Accelerated Diffusion Transformers," Accepted at CVPR, 2026 (*co-first authors) (Highlight)
  - Seongmin Hong* (Postdoc), Junghun J. Kim* (MS student), Daehyeop Kim (Undergrad intern), Insoo Chung (Undergrad intern), Se Young Chun, "DiffBMP: Differentiable Rendering with Bitmap Primitives," Accepted at CVPR, 2026 (*co-first authors)
  - Jinho Park, Se Young Chun, Mingoo Seok, "AdaRadar: Rate Adaptive Spectral Compression for Radar-based Perception," Accepted at CVPR, 2026
(2026 Mar) One collaboration paper published at Theranostics (IF 13.3)
- - S Y Jo, J W Shon, S H An, Y J Nam, D W Choi, S M Lee, Hoigi Seo (MS/PhD student), Se Young Chun, H Kim, G H Kim, “Tailorable porous collagen hydrogels as a physiologically relevant platform for extrachromosomal DNA-associated colorectal cancer research,” Theranostics 16(10):5240-58, Mar 2026
(2026 Mar) Our lab (ICL@SNU), in collaboration with Prof. Hoon Kim at SKKU, has been selected for the NVIDIA Academic Grant Program. NVIDIA will be providing 32,000 A100 GPU-hours to SNU in support of our megabase-scale DNA foundation model project.
(2026 Mar) One paper accepted at IEEE Access (IEEE Geoscience and Remote Sensing Society Section) (IF: 3.6).
- - Sungjin Lim (MS student)*, Dong Un Kang* (PhD graduate)*, Yeong Mok Cho (Undergrad intern), Se Young Chun, "Dual Context-Conditioned Open-Vocabulary Semantic Segmentation in Remote Sensing," Accepted at IEEE Access (IEEE Geoscience and Remote Sensing Society Section), Mar 2026 (*co-first authors)
(2026 Feb) Sungjin Lim won the IPAI Best Master's Thesis Award. Congratulations!
(2026 Jan) Dohyun Mah (ECE@SNU) and Chaeeun Lee (AAI@SKKU) will join our Lab as MS-PhD students. Daewon Choi (Samsung DS, EEE@Yonsei) will join our Lab as a MS student.
(2025 Dec) Joonhee Lee and Gwanghyun Kim have successfully completed PhD thesis defense. Sungjin Lim has successfully completed MS thesis defense. Congratulations!
(2025 Oct) One paper accepted at Image and Vision Computing (IF 4.2) and two collaboration papers accepted at IEEE Access (IF 3.6) and IEEE Transactions on Computation Imaging (IF 4.8)
- - Haechang Lee* (PhD student), Dong Ju Mun* (MS/PhD student), Hyunwoo Lee* (MS student), Kyung Chul Lee, Seongmin Hong (Postdoc), Gwanghyun Kim (PhD student), Seung Ah Lee**, Se Young Chun**, "Fidelity-Preserving Zero-shot Diffusion Models for Highly Ill-posed Inverse Problems in Lensless Imaging," Accepted at Image and Vision Computing, 2025 (*co-first authors, **co-corresponding authors)
  - J H Yoon, H G Park, M J Kim, H H Jeong, Se Young Chun, S Y Ji, J H Lee, "Clinical dementia rating classification using integrated vision and language information," Accepted at IEEE Access, 2025
  - T Hong, Z Xu, Se Young Chun, L Hernandez-Garcia, J A Fessler, “Convergent Complex Quasi-Newton Proximal Methods for Gradient-Driven Denoisers in Compressed Sensing MRI Reconstruction,” Accepted at IEEE Transactions on Computational Imaging, 2025
(2025 Oct) Byung Hyun Lee has joined Naver Cloud as an intern.
(2025 Oct) Hoigi Seo won a BK21Four - SNU Fellowship of Research Excellence. Congratulations!
(2025 Sep) ICL won the 2nd place at MICCAI'25 Challenge on REport Generation in pathology using Pan-Asia Giga-pixel WSIs (REG). Congratulations, Byung Hyun Lee, Sungjin Lim, Joohoon Lee, Gwanghyun Kim!

Page updated

Google Sites

Report abuse