Adapting Diffusion Policies to Novel Environments via Policy-Steered Optimization

Skye Thompson, Sergio Orozco, Eric Rosen, Karl Schmeckpeper, George Konidaris

Keywords: robotics, policy steering, optimization

TL;DR: Adapting diffusion policies to novel environment constraints by using policy-aligment as an optimization objective.

Abstract: We propose a novel method of adapting generative robot control policies to succeed in unfamiliar environments with novel runtime constraints. We use a model-based planner to optimize for trajectories that obey both the novel environmental constraints and the implicit task constraints learned by the policy model. We achieve this by evaluating a policy-alignment objective that measures the policy model’s success at reconstructing noised trajectories. We demonstrate this approach’s ability to generalize to two novel simulation environments with obstacles not seen during training.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 54