Publications
International Conference
- Jung Min Lee, Dohyeok Lee, Jungwoo Lee, ****Viewpoint-Invariant Latent Action Learning from Human Video Demonstrations, NeurIPS SpaVLE Workshop, 2025.
- Dohyeok Lee, Jung Min Lee, Munkyung Kim, Seokhun Ju, Seungyub Han, Jin Woo Koo, Jungwoo Lee, Dynamics-Aligned Flow Matching Policy for Robot Learning, CVPR EAI Workshop, 2025.
- Dohyeok Lee, Munkyung Kim, Jung Min Lee, Seungyub Han, Jungwoo Lee, View-Imagination: Enhancing Visuomotor Control with Adaptive View
Synthesis, CVPR EAI Workshop, 2025.
Preprint
- Dohyeok Lee, Jung Min Lee, Munkyung Kim, Seokhun Ju, Jin Woo Koo, Kyungjae Lee, Dohyeong Kim, TaeHyun Cho, Jungwoo Lee, Learning Generalizable Visuomotor Policy through Dynamics-Alignment, arXiv preprint, 2025.
Manuscript in preparation
- Understanding the locomotion of Hexapod organism via Tegotae model - in terms of nonlinear dynamics
- Representation Learning from Human video demonstration
Open Source Contributions
- A custom Reinforcement Learning (RL) environment using the Franka Panda arm
- Visoumotor control : Soft Actor-Critic(SAC) agent with single RGB image as an observation
- 4 Manipulation tasks : Reach, Push, Lift, and Pick-and-place
Honor