UniSkill: Imitating Human Videos via Cross-Embodiment Skill Representations

TL;DR

UniSkill introduces a framework for learning embodiment-agnostic skill representations from large-scale cross-embodiment videos, enabling effective transfer of human skills to robots without labeled data, demonstrated in simulation and real-world tests.

Contribution

It presents a novel, label-free method for extracting transferable, embodiment-agnostic skills from videos, bridging the gap between human demonstrations and robot execution.

Findings

Skills successfully transferred from human videos to robots in various tasks.

Robots can follow unseen human video prompts to perform tasks.

Effective in both simulation and real-world environments.

Abstract

Mimicry is a fundamental learning mechanism in humans, enabling individuals to learn new tasks by observing and imitating experts. However, applying this ability to robots presents significant challenges due to the inherent differences between human and robot embodiments in both their visual appearance and physical capabilities. While previous methods bridge this gap using cross-embodiment datasets with shared scenes and tasks, collecting such aligned data between humans and robots at scale is not trivial. In this paper, we propose UniSkill, a novel framework that learns embodiment-agnostic skill representations from large-scale cross-embodiment video data without any labels, enabling skills extracted from human video prompts to effectively transfer to robot policies trained only on robot data. Our experiments in both simulation and real-world environments show that our cross-embodiment skills successfully guide robots in selecting appropriate actions, even with unseen video prompts. The project website can be found at: https://kimhanjung.github.io/UniSkill.