Hello, I'm Juhan Park

I am a Ph.D student at Korea Unversity, where I work with Sungjoon Choi on artificial inteligence and robotics. Previously, I received my M.S. from the Department of Artificial Intelligence at Chung-Ang University, advised by Kyungjae Lee.

My focus is on robot learning for autonomous systems, with a recent emphasis on dexterous hand manipulation. I aspire to integrate this robotic intelligence into the agricultural sector to enable sophisticated automation.

Scholar Github Linkedin

News

[June 2026]: Our paper about Learning Dexterous Grasping from Sparse Taxonomy Guidance got accepted to IROS 2026.
[October 2025]: Our lab is conducting a joint demonstration with RLWRLD. (video link)
[March 2025]: Started my Ph.D. at Korea University, advised by Sungjoon Choi.
[February 2025]: Graduated from the Department of Artificial Intelligence at Chung-Ang University with a M.S. degree..
[Feburary 2025]: I am presenting our work about placement aware grasp planning in Korea Robotics Society Annual Conference (KRoC) as invited speaker in flagship conference session.
[October 2024]: Our paper about Placement Aware Grasp Planning for Efficient Sequential Manipulation got published to ECAI 2024.
[May 2023]: Our paper about Efficient Task Planning with MCTS got published to ICRA 2023.
[March 2023]: I have started my M.S. at the Department of Artificial Intelligence at Chung-Ang University, advised by Kyungjae Lee.

Publications

Retrieve, Don't Retrain: Extending Vision-Language-Action Models to New Tasks at Test Time

Jeongeun Park, Juhan Park, Taekyung Kim, Sungjoon Choi, Dongyoon Han, Sangdoo Yun

Preprint, 2026

ReCAP is a retrieval-augmented policy that adapts vision-language-action models to new tasks at deployment without retraining: the frozen policy conditions on retrieved trajectories at every control step, so new tasks are absorbed by indexing data rather than updating parameters.

Project Page

Learning Dexterous Grasping from Sparse Taxonomy Guidance

Juhan Park, Taerim Yoon, Seungmin Kim, Joonggil Kim, Wontae Ye, Jeongeun Park, Yoonbyung Chai, Geonwoo Cho, Geunwoo Cho, Dohyeong Kim, Kyungjae Lee, Yongjae Kim, Sungjoon Choi

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2026

Dexterous Grasping from Sparse Taxonomy Guidance.

Project Page arXiv Video

Hierarchical Vision Language Action Model Using Success and Failure Demonstrations

Jeongeun Park, Jihwan Yoon, Byeongwoo Jeon, Juhan Park, Jinwoo Shin, Namhoon Cho, Kyungjae Lee, Sangdoo Yun, Sungjoon Choi

CoRL 2025 2nd Workshop on Safe and Robust Robot Learning for Operation in the Real World

We propose VINE, a dual-system framework that injects failure-aware reasoning into VLAs. System 1 executes grounded action chunks, while System 2 builds a tree of thought states and scores candidate subgoals using both success and failure data.

Project Page arXiv