General and Efficient Visual Goal-Conditioned Reinforcement Learning using Object-Agnostic Masks

Open in new window