I am Seunghun Lee, a Ph.D. student at DGIST (Daegu Gyeongbuk Institute of Science and Technology, Daegu, Republic of Korea) under the supervision of Prof. Sunghoon Im. I received my B.S. degree in Convergence Engineering from DGIST. My previous research focused on domain adaptation using image-to-image translation. Currently, my research interests lie in object-centric video tasks such as Video Instance Segmentation (VIS) and Referring Video Object Segmentation (RVOS). More recently, I have been expanding my research toward video reasoning tasks using multi-modal large language models (MLLM). From September 2024 to February 2025, I worked as a visiting researcher at Stanford University with Prof. Ehsan Adeli.

💡 I am currently seeking job opportunities, so please feel free to reach out if you are interested.

🔥 News

2025.09: 🏆 Our team achieved 1st place in the 4th GroceryVision Challenge on Spatial Temporal Action Localization at ICCV 2025.
2025.09: 🎉 Two papers accepted to ICCVw 2025.
2025.06: 🎉 Two papers accepted to ICCV 2025.
2025.02: 🎉 One paper accepted to CVPR 2025.

📝 Publications

🚀 Recent Works

SAMDWICH: Moment-aware Video-text Alignment for Referring Video Object Segmentation

Seunghun Lee*, Jiwan Seo*, Jeonghoon Kim*, Siwon Kim, Haeun Yun, Hyogyeong Jeon, Wonhyeok Choi, Jaehoon Jeong, Zane Durante, Sang Hyun Park, Sunghoon Im†

Paper | Project | Code

ICCV 2025

Latest Object Memory Management for Temporally Consistent Video Instance Segmentation

Seunghun Lee, Jiwan Seo, Minwoo Choi, Kiljoon Han, Jaehoon Jeong, Zane Durante, Ehsan Adeli†, Sang Hyun Park, Sunghoon Im†

Paper | Project | Code

ICCV 2025

CAVIS: Context-Aware Video Instance Segmentation

Seunghun Lee*, Jiwan Seo*, Kiljoon Han, Minwoo Choi, Sunghoon Im†

Paper | Project | Code |

CVPR 2025

Style-Editor: Text-driven object-centric style editing (Highlight, Top 3.7%)

Jihun Park*, Jongmin Gim*, Kyoungmin Lee*, Seunghun Lee, Sunghoon Im†

Paper | Project

📚 Others

WACV 2024 (Oral, Top 2.6%) Offline-to-Online Knowledge Distillation for Video Instance Segmentation, Hojin Kim, Seunghun Lee, Hyeon Kang, Sunghoon Im†
BMVC 2023 Multi-Target Domain Adaptation with Class-Wise Attribute Transfer in Semantic Segmentation, Changjae Kim, Seunghun Lee, Sunghoon Im†
CVPR 2022 ADAS: A Direct Adaptation Strategy for Multi-Target Domain Adaptive Semantic Segmentation, Seunghun Lee, Wonhyeok Choi, Changjae Kim, Minwoo Choi, Sunghoon Im† | Code
CVPR 2021 DRANet: Disentangling Representation and Adaptation Networks for Unsupervised Cross-Domain Adaptation, Seunghun Lee, Sunghyun Cho, Sunghoon Im† | Code

🎖 Honors and Awards

2025.09, 1st Place – 4th GroceryVision Challenge (Spatial-Temporal Action Localization), ICCV 2025
2024.07, DGIST Post-Graduate Research Abroad Awards, DGIST
2024.02, Participation prize, 30th HumanTech Paper Award, Samsung Electronics Co., Ltd.
2022.08, Outstanding Researcher Award, 2022 DGIST Student Conference, EECS.
2022.02, Participation prize, 28th HumanTech Paper Award, Samsung Electronics Co., Ltd.
2021.12, 3rd place Prize at ICT Paper Contest.

🏫 Educations

2019.02 - 2026.02 (Expected), M.S-Ph.D. Course, Electrical Engineering and Computer Science, DGIST, Daegu.
2015.02 - 2019.02, Undergraduate, Convergence Engineering, DGIST, Daegu.

📖 Academic Activities

Reviewer
- 2023 - present, IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- 2023 - present, Conference on Neural Information Processing Systems (NeurIPS)
- 2023 - present, IEEE International Conference on Computer Vision (ICCV)
- 2024 - present, International Conference on Machine Learning (ICML)
- 2024 - present, European Conference on Computer Vision (ECCV)
- 2024 - present, Association for the Advancement of Artificial Intelligence (AAAI)
- 2024 - present, International Conference on Learning Representations (ICLR)
- 2024, Pacific Graphics
Advanced Studies
- 2023, International Computer Vision Summer School (ICVSS)

💻 Research Experience

2024.09 - 2025.02, Visiting Researcher, Stanford University, California, US.