Oral

Self-Imitation Learning via Trajectory-Conditioned Policy for Hard-Exploration Tasks

Unsupervised Discovery of Object Landmarks as Structural Representations