Data Efficient Behavior Cloning for Fine Manipulation via Continuity-based Corrective Labels

TL;DR

This paper introduces CCIL, a novel imitation learning framework that uses continuity-based corrective labels to improve fine manipulation tasks on robots, effectively reducing errors caused by covariate shift.

Contribution

The paper presents CCIL, a new method that leverages local smoothness in contact-rich manipulation to generate corrective labels, enhancing imitation learning performance in real-world robotic tasks.

Findings

CCIL significantly improves imitation learning in contact-rich manipulation.

Local smoothness in real-world tasks enables effective use of CCIL.

Corrective labels are most beneficial in low-data regimes.

Abstract

We consider imitation learning with access only to expert demonstrations, whose real-world application is often limited by covariate shift due to compounding errors during execution. We investigate the effectiveness of the Continuity-based Corrective Labels for Imitation Learning (CCIL) framework in mitigating this issue for real-world fine manipulation tasks. CCIL generates corrective labels by learning a locally continuous dynamics model from demonstrations to guide the agent back toward expert states. Through extensive experiments on peg insertion and fine grasping, we provide the first empirical validation that CCIL can significantly improve imitation learning performance despite discontinuities present in contact-rich manipulation. We find that: (1) real-world manipulation exhibits sufficient local smoothness to apply CCIL, (2) generated corrective labels are most beneficial in low-data regimes, and (3) label filtering based on estimated dynamics model error enables performance gains. To effectively apply CCIL to robotic domains, we offer a practical instantiation of the framework and insights into design choices and hyperparameter selection. Our work demonstrates CCIL's practicality for alleviating compounding errors in imitation learning on physical robots.