SkillPlug: Unsupervised Skill Mining for Few-Shot Adaptation in Robotic Manipulation

2026-07-09

AI Summary

This work addresses the limited cross-task generalization and inefficient few-shot adaptation of existing end-to-end robotic manipulation policies, which stem from a lack of reusable behavioral structures. To overcome this, the authors propose SkillPlug, a framework that leverages self-supervised objectives to unsupervisely extract compact, non-redundant, and transferable skill primitives from raw multi-task demonstrations, thereby constructing a shared skill prior across tasks. SkillPlug further introduces a skill-conditioned policy architecture featuring lightweight routing and action heads, enabling plug-and-play fine-tuning without end-to-end retraining. Experiments demonstrate that SkillPlug significantly enhances multi-task performance and few-shot adaptation on two simulated benchmarks and a real robot, achieving more efficient data utilization and improved generalization.

Abstract

Learning transferable visuomotor imitation policies that generalize across diverse manipulation tasks and adapt rapidly to new tasks from only a handful of demonstrations remains challenging. Most modern policies are trained end-to-end to map observations directly to low-level actions, offering little explicit structure for reusing and recombining behaviors across tasks and making transfer data-inefficient under limited supervision. We propose SkillPlug, a plug-in framework that augments an existing visuomotor policy with a skill-conditioning module and mines a shared, transferable skill library from raw multi-task demonstrations. SkillPlug learns skills via self-supervised objectives that promote compact, reusable, and non-redundant behavior-level primitives, forming a task-shared prior for compositional control. After skill mining, we keep the learned skills fixed and specialize to unseen tasks by fine-tuning only lightweight router and action head, enabling efficient adaptation without full end-to-end retraining. We evaluate SkillPlug on two simulation benchmarks and on a real robot, and observe that the mined transferable skills consistently improve both multi-task performance and few-shot adaptation. Overall, SkillPlug offers a scalable way to mine reusable skills that improve data-efficient generalization in robotic manipulation.

Problem

• few-shot adaptation
• skill transfer
• robotic manipulation
• visuomotor imitation
• data-efficient learning

Innovation

• skill mining
• few-shot adaptation
• visuomotor imitation
• self-supervised learning
• compositional control