R+X: Retrieval and Execution from Everyday Human Videos

TL;DR

R+X is a novel framework that enables robots to learn and execute household skills directly from unlabelled human videos by combining retrieval via vision-language models and in-context imitation learning, eliminating manual annotations and training phases.

Contribution

It introduces a retrieval and execution framework that learns robot skills from unlabelled human videos using vision-language models and in-context learning, without manual annotations or training.

Findings

R+X successfully translates unlabelled human videos into robot skills.

R+X outperforms recent alternative methods in household tasks.

The framework enables immediate skill execution without training.

Abstract

We present R+X, a framework which enables robots to learn skills from long, unlabelled, first-person videos of humans performing everyday tasks. Given a language command from a human, R+X first retrieves short video clips containing relevant behaviour, and then executes the skill by conditioning an in-context imitation learning method (KAT) on this behaviour. By leveraging a Vision Language Model (VLM) for retrieval, R+X does not require any manual annotation of the videos, and by leveraging in-context learning for execution, robots can perform commanded skills immediately, without requiring a period of training on the retrieved videos. Experiments studying a range of everyday household tasks show that R+X succeeds at translating unlabelled human videos into robust robot skills, and that R+X outperforms several recent alternative methods. Videos and code are available at https://www.robot-learning.uk/r-plus-x.