One-Shot Cross-Geometry Skill Transfer through Part Decomposition

2026-04-16

AI Summary

Existing skill transfer methods struggle to generalize to novel objects with unfamiliar geometries and fail to enable effective manipulation from a single demonstration. This work proposes a novel approach that integrates semantic part decomposition with a data-efficient generative shape model to accurately transfer interaction points from a demonstration to the corresponding semantic parts of a new object and automatically construct an optimization objective that aligns skill-relevant components. For the first time, this method achieves one-shot skill transfer across diverse geometric morphologies, successfully transferring multiple manipulation skills in both simulation and real-world settings using only a single demonstration. It substantially outperforms existing approaches and significantly enhances generalization to previously unseen objects.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

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.

Problem

• skill transfer
• cross-geometry
• one-shot learning
• object generalization
• shape adaptation

Innovation

• part decomposition
• one-shot skill transfer
• generative shape models
• cross-geometry generalization
• semantic parts