Scaling Robot Policy Learning via Zero-Shot Labeling with Foundation Models

TL;DR

This paper introduces NILS, a zero-shot, automatic labeling method using foundation models to scale robot policy learning by annotating large, unstructured datasets without human input.

Contribution

NILS is a novel approach that leverages pretrained vision-language models to automatically label robot data at scale, improving diversity and quality over traditional human annotations.

Findings

NILS successfully labeled over 115,000 robot trajectories.

It improved annotation diversity and quality compared to crowdsourced labels.

The method enables scalable, zero-shot annotation of unstructured robot datasets.

Abstract

A central challenge towards developing robots that can relate human language to their perception and actions is the scarcity of natural language annotations in diverse robot datasets. Moreover, robot policies that follow natural language instructions are typically trained on either templated language or expensive human-labeled instructions, hindering their scalability. To this end, we introduce NILS: Natural language Instruction Labeling for Scalability. NILS automatically labels uncurated, long-horizon robot data at scale in a zero-shot manner without any human intervention. NILS combines pretrained vision-language foundation models in order to detect objects in a scene, detect object-centric changes, segment tasks from large datasets of unlabelled interaction data and ultimately label behavior datasets. Evaluations on BridgeV2, Fractal, and a kitchen play dataset show that NILS can autonomously annotate diverse robot demonstrations of unlabeled and unstructured datasets while alleviating several shortcomings of crowdsourced human annotations, such as low data quality and diversity. We use NILS to label over 115k trajectories obtained from over 430 hours of robot data. We open-source our auto-labeling code and generated annotations on our website: http://robottasklabeling.github.io.