- Dongguk University, Seoul (Department of Computer Engineering)
- Human-Robot alignment
- EgoBridge: Domain Adaptation for Generalizable Imitation from Egocentric Human Data
- EgoVLA: Learning Vision-Language-Action Models from Egocentric Human Videos
- Domain adaptation
- Unsupervised Domain Adaptation for Ship Classification via Progressive Feature Alignment: From Optical to SAR Images
- Representation Learning
- MedCLIP-SAMv2: Towards Universal Text-Driven Medical Image Segmentation
- Latent Action Pretraining from Videos
- Others
- Direct Preference Optimization: Your Language Model is Secretly a Reward Model
