Fetching the paper…

Self-supervised Visual Reinforcement Learning with Object-centric Representations · Around