Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hao, Ce, Zhai, Xuanran, Liu, Yaohua, Soh, Harold
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914289458085888
author Hao, Ce
Zhai, Xuanran
Liu, Yaohua
Soh, Harold
author_facet Hao, Ce
Zhai, Xuanran
Liu, Yaohua
Soh, Harold
contents Diffusion-based policies have recently shown strong results in robot manipulation, but their extension to multi-task scenarios is hindered by the high cost of scaling model size and demonstrations. We introduce Skill Mixture-of-Experts Policy (SMP), a diffusion-based mixture-of-experts policy that learns a compact orthogonal skill basis and uses sticky routing to compose actions from a small, task-relevant subset of experts at each step. A variational training objective supports this design, and adaptive expert activation at inference yields fast sampling without oversized backbones. We validate SMP in simulation and on a real dual-arm platform with multi-task learning and transfer learning tasks, where SMP achieves higher success rates and markedly lower inference cost than large diffusion baselines. These results indicate a practical path toward scalable, transferable multi-task manipulation: learn reusable skills once, activate only what is needed, and adapt quickly when tasks change.
format Preprint
id arxiv_https___arxiv_org_abs_2601_21251
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies
Hao, Ce
Zhai, Xuanran
Liu, Yaohua
Soh, Harold
Robotics
Diffusion-based policies have recently shown strong results in robot manipulation, but their extension to multi-task scenarios is hindered by the high cost of scaling model size and demonstrations. We introduce Skill Mixture-of-Experts Policy (SMP), a diffusion-based mixture-of-experts policy that learns a compact orthogonal skill basis and uses sticky routing to compose actions from a small, task-relevant subset of experts at each step. A variational training objective supports this design, and adaptive expert activation at inference yields fast sampling without oversized backbones. We validate SMP in simulation and on a real dual-arm platform with multi-task learning and transfer learning tasks, where SMP achieves higher success rates and markedly lower inference cost than large diffusion baselines. These results indicate a practical path toward scalable, transferable multi-task manipulation: learn reusable skills once, activate only what is needed, and adapt quickly when tasks change.
title Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies
topic Robotics
url https://arxiv.org/abs/2601.21251