Yeeun Choi

PhD Student @ Yonsei University | Video Understanding

profile.jpg

I am a Ph.D. student in Artificial Intelligence at Yonsei University, advised by Prof. Seon Joo Kim in the Computational Intelligence & Photography Lab (CIPLAB).

My research centers on video understanding and multi-modal large language models (MLLMs). My long-term goal is to build a real-life AI assistant that perceives and reasons over what we see and hear in everyday life. Toward this goal, I work on ultra-long video understanding, with a particular focus on egocentric video, as well as streaming video understanding that processes continuous video in real time. I am also interested in integrating diverse modalities beyond vision and language to build more context-aware and adaptive intelligent systems.

News

Aug. 2026 MERIT is selected as a Long Oral presentation at ECCV 2026! :tada:
Jun. 2026 MERIT is accepted to ECCV 2026. See you in Malmö! :sweden:
Jun. 2025 OpenHouse is accepted to ICCV 2025.
Apr. 2025 ECP is accepted to CVPR 2025 EVAL-FoMo-2 Workshop.

Publications

(* indicates equal contributions)

  1. Keep It Simple: Multi-Key Episodic Memory Retrieval for Ultra-Long Video Understanding thumbnail
    Keep It Simple: Multi-Key Episodic Memory Retrieval for Ultra-Long Video Understanding
    ECCV 2026 (Oral)
  2. Open-ended Hierarchical Streaming Video Understanding with Vision Language Models thumbnail
    Open-ended Hierarchical Streaming Video Understanding with Vision Language Models
    ICCV 2025
  3. A Training-Free, Task-Agnostic Framework for Enhancing MLLM Performance on High-Resolution Images thumbnail
    A Training-Free, Task-Agnostic Framework for Enhancing MLLM Performance on High-Resolution Images
    Jaeseong Lee* , Yeeun Choi* , Heechan Choi* , Hanjung Kim , Seon Joo Kim
    CVPRW 2025