Online Imitation Learning for Manipulation via Decaying Relative Correction through Teleoperation

TL;DR

This paper introduces Decaying Relative Correction (DRC), a method that uses spatial offsets from experts to improve teleoperation-based imitation learning, reducing expert intervention and enhancing manipulation success rates.

Contribution

The paper presents DRC, a novel correction technique that decreases expert intervention by 30% and accelerates learning in manipulation tasks within an online imitation learning framework.

Findings

DRC reduces expert intervention rate by 30%.

Integrating DRC improves success rates in manipulation tasks.

The method enables rapid policy adaptation with less expert effort.

Abstract

Teleoperated robotic manipulators enable the collection of demonstration data, which can be used to train control policies through imitation learning. However, such methods can require significant amounts of training data to develop robust policies or adapt them to new and unseen tasks. While expert feedback can significantly enhance policy performance, providing continuous feedback can be cognitively demanding and time-consuming for experts. To address this challenge, we propose to use a cable-driven teleoperation system which can provide spatial corrections with 6 degree of freedom to the trajectories generated by a policy model. Specifically, we propose a correction method termed Decaying Relative Correction (DRC) which is based upon the spatial offset vector provided by the expert and exists temporarily, and which reduces the intervention steps required by an expert. Our results demonstrate that DRC reduces the required expert intervention rate by 30\% compared to a standard absolute corrective method. Furthermore, we show that integrating DRC within an online imitation learning framework rapidly increases the success rate of manipulation tasks such as raspberry harvesting and cloth wiping.