|国家预印本平台
首页|AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation

AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation

AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation

来源:Arxiv_logoArxiv
英文摘要

Generalizing control policies to novel embodiments remains a fundamental challenge in enabling scalable and transferable learning in robotics. While prior works have explored this in locomotion, a systematic study in the context of manipulation tasks remains limited, partly due to the lack of standardized benchmarks. In this paper, we introduce a benchmark for learning cross-embodiment manipulation, focusing on two foundational tasks-reach and push-across a diverse range of morphologies. The benchmark is designed to test generalization along three axes: interpolation (testing performance within a robot category that shares the same link structure), extrapolation (testing on a robot with a different link structure), and composition (testing on combinations of link structures). On the benchmark, we evaluate the ability of different RL policies to learn from multiple morphologies and to generalize to novel ones. Our study aims to answer whether morphology-aware training can outperform single-embodiment baselines, whether zero-shot generalization to unseen morphologies is feasible, and how consistently these patterns hold across different generalization regimes. The results highlight the current limitations of multi-embodiment learning and provide insights into how architectural and training design choices influence policy generalization.

Meenal Parakh、Alexandre Kirchmeyer、Beining Han、Jia Deng

自动化基础理论计算技术、计算机技术

Meenal Parakh,Alexandre Kirchmeyer,Beining Han,Jia Deng.AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation[EB/OL].(2025-05-20)[2025-08-02].https://arxiv.org/abs/2505.14986.点此复制

评论