Reinforcement Learning with Neural Radiance Fields

Danny Driess, Ingmar Schubert, Pete Florence +2

It is a long-standing problem to find effective representations for training reinforcement learning (RL) agents. This paper demonstrates that learning state representations with su…