Published multiple papers at top-tier venues including NeurIPS 2025, ICCV 2025, CVPR 2025 (oral), AAAI 2025, TPAMI 2024, ECCV 2024, CVPR 2023, AAAI 2023 (oral), ECCV 2022, and CVPR 2022 (oral)
Research topics span partially relevant video retrieval, few-shot action recognition, unsupervised semantic segmentation, video temporal grounding, and zero-shot object counting
Also published at domestic conferences KIISE 2022 and KIISE 2020
Several papers list him as co-first author (marked with *)
Many works include open-source code and video demos
Background
Ph.D. student at Sungkyunkwan University, South Korea
Member of Visual Computing Laboratory (VCLAB), advised by Prof. Jae-Pil Heo
Primary research interests: Video Understanding and Multi-modal Learning
Recently focused on Video Object-Centric Learning for efficient video representations
Other interests include Video Retrieval, Representation Learning, Segmentation, and Few-Shot Learning