Dreamitate: Real-World Visuomotor Policy Learning via Video Generation

TL;DR

Dreamitate introduces a visuomotor policy learning method that fine-tunes video diffusion models on human demonstrations, enabling robots to generalize across diverse visual environments by generating task executions conditioned on novel scenes.

Contribution

The paper presents a novel framework that leverages pretrained video generative models for robot policy learning, bridging the embodiment gap using common tools and demonstrating superior generalization.

Findings

Outperforms existing behavior cloning methods in generalization.

Successfully applies to four tasks of increasing complexity.

Utilizes internet-scale video models for robust visuomotor control.

Abstract

A key challenge in manipulation is learning a policy that can robustly generalize to diverse visual environments. A promising mechanism for learning robust policies is to leverage video generative models, which are pretrained on large-scale datasets of internet videos. In this paper, we propose a visuomotor policy learning framework that fine-tunes a video diffusion model on human demonstrations of a given task. At test time, we generate an example of an execution of the task conditioned on images of a novel scene, and use this synthesized execution directly to control the robot. Our key insight is that using common tools allows us to effortlessly bridge the embodiment gap between the human hand and the robot manipulator. We evaluate our approach on four tasks of increasing complexity and demonstrate that harnessing internet-scale generative models allows the learned policy to achieve a significantly higher degree of generalization than existing behavior cloning approaches.