Motion Before Action: Diffusing Object Motion as Manipulation Condition

TL;DR

This paper presents MBA, a diffusion-based module that predicts object motion to guide robotic manipulation, significantly enhancing policy performance in various tasks.

Contribution

Introduces MBA, a novel diffusion-based framework that models object motion to improve robot manipulation, with flexible integration into existing policies.

Findings

Substantially improves manipulation performance in simulations.

Effective in real-world robotic tasks.

Versatile and plug-and-play design.

Abstract

Inferring object motion representations from observations enhances the performance of robotic manipulation tasks. This paper introduces a new paradigm for robot imitation learning that generates action sequences by reasoning about object motion from visual observations. We propose MBA (Motion Before Action), a novel module that employs two cascaded diffusion processes for object motion generation and robot action generation under object motion guidance. MBA first predicts the future pose sequence of the object based on observations, then uses this sequence as a condition to guide robot action generation. Designed as a plug-and-play component, MBA can be flexibly integrated into existing robotic manipulation policies with diffusion action heads. Extensive experiments in both simulated and real-world environments demonstrate that our approach substantially improves the performance of existing policies across a wide range of manipulation tasks. Project page: https://selen-suyue.github.io/MBApage/