ObstaDiff

Generalizable Diffusion Policy Learning via Obstacle-aware Representations