One-Shot Cross-Geometry Skill Transfer through Part Decomposition

TL;DR

This paper introduces a method for robot skill transfer that decomposes objects into parts, enabling one-shot generalization to new objects with different shapes using shape models and part alignment.

Contribution

It proposes a novel part decomposition approach combined with generative shape models for effective one-shot skill transfer to unfamiliar object geometries.

Findings

Achieves successful one-shot transfer for various skills and objects from a single demonstration.

Generalizes to a wider range of object geometries than previous methods.

Works effectively in both simulated and real environments.

Abstract

Given a demonstration, a robot should be able to generalize a skill to any object it encounters-but existing approaches to skill transfer often fail to adapt to objects with unfamiliar shapes. Motivated by examples of improved transfer from compositional modeling, we propose a method for improving transfer by decomposing objects into their constituent semantic parts. We leverage data-efficient generative shape models to accurately transfer interaction points from the parts of a demonstration object to a novel object. We autonomously construct an objective to optimize the alignment of those points on skill-relevant object parts. Our method generalizes to a wider range of object geometries than existing work, and achieves successful one-shot transfer for a range of skills and objects from a single demonstration, in both simulated and real environments.