JunGyu Lee
I'm a Ph.D. student at the Graduate School of Artificial Intelligence, Yonsei University, where I'm advised by Hae-Gon Jeon in the Visual AI Lab. My research interests include crowd generation, social robot navigation, and world models.
Previously, I received my M.S. in AI Robotics from KIST School (UST), advised by Gi Pyo Nam, and worked as an AI researcher at KIST and Seoul National University Hospital.
My goal is to develop human-centric AI that addresses complex societal challenges and creates meaningful real-world impact.
News
- Oct 2026Our preprint Ctrl-CWM on controllable crowd generation through world-model planning is now on arXiv, with code.
- Oct 2026Our preprint MoRE on language-based trajectory prediction is now on arXiv.
- Oct 2026Our preprint IORA on support-aware grasp synthesis for immobilizing irregular objects with low-DoF robot hands is now on arXiv.
- Oct 2026Our preprint SGP-TTA on test-time adaptation for thin curvilinear structure segmentation is now on arXiv, with code.
- Sep 2026ComPose is accepted to NeurIPS 2026.
- Jul 2026PSCT-Net is accepted to the MICCAI 2026 Workshop.
- Jun 2026DTC-TRUS is accepted to MICCAI 2026.
- Mar 2026Started my Ph.D. at Yonsei University in the Visual AI Lab, advised by Hae-Gon Jeon.
- Mar 2025V-NAW is accepted to the CVPR 2025 Workshop, with a Top 5 placement at the 8th ABAW Competition.
- Dec 2024NLA is accepted to AAAI 2025.
Research
I'm interested in crowd generation, social robot navigation, and world models, and have also worked on facial expression recognition and medical image analysis.
* equal contribution
Experience
- AI Researcher Seoul National University Hospital (SNUH)
- AI Researcher Korea Institute of Science and Technology (KIST)
- Research Intern Asan Medical Center (AMC)
Education
- Ph.D. in Engineering, Artificial Intelligence Yonsei University
- Master of Engineering, AI Robotics University of Science and Technology (UST), KIST School
- B.S. in Big Data Analytics Kookmin University
Honors & Awards
- 2025Top 5 Placement, 8th ABAW Competition @ CVPR Workshop
- 2024Best Paper Award, KIST Convergence Conference
- 2024Outstanding Technology Award, AI-Robotics Research Division, KIST