From Pixels to Factors: Learning Independently Controllable State Variables for Reinforcement Learning

Open in new window