Refactoring Policy for Compositional Generalizability using Self-Supervised Object Proposals

Understand

We study how to learn a policy with compositional generalizability.

• We propose a two-stage framework, which refactorizes a high-reward teacher policy into a generalizable student policy with strong inductive bias.
• Particularly, we implement an object-centric GNN-based student policy, whose input objects are learned from images through self-supervised learning.
• Empirically, we evaluate our approach on four difficult tasks that require compositional generalizability, and achieve superior performance compared to baselines.