MoVie: Visual Model-Based Policy Adaptation for View Generalization Sizhe Y ang

Visual Reinforcement Learning (RL) agents trained on limited views face significant challenges in generalizing their learned abilities to unseen views.