Rank2Reward: Learning Shaped Reward Functions from Passive Video

Open in new window