One-Shot Domain-Adaptive Imitation Learning via Progressive Learning

TL;DR

This paper introduces a one-shot domain-adaptive imitation learning framework that employs progressive learning phases to enable robots to adapt to new scenarios efficiently with minimal demonstration data.

Contribution

It proposes a novel three-phase progressive learning approach for one-shot domain adaptation in imitation learning, improving success rates and generalizability over existing methods.

Findings

Enhanced success rate in robotic pouring tasks

More efficient training process

Better generalization to new domains with diverse backgrounds and objects

Abstract

Traditional deep learning-based visual imitation learning techniques require a large amount of demonstration data for model training, and the pre-trained models are difficult to adapt to new scenarios. To address these limitations, we propose a unified framework using a novel progressive learning approach comprised of three phases: i) a coarse learning phase for concept representation, ii) a fine learning phase for action generation, and iii) an imaginary learning phase for domain adaptation. Overall, this approach leads to a one-shot domain-adaptive imitation learning framework. We use robotic pouring task as an example to evaluate its effectiveness. Our results show that the method has several advantages over contemporary end-to-end imitation learning approaches, including an improved success rate for task execution and more efficient training for deep imitation learning. In addition, the generalizability to new domains is improved, as demonstrated here with novel background, target container and granule combinations. We believe that the proposed method can be broadly applicable to different industrial or domestic applications that involve deep imitation learning for robotic manipulation, where the target scenarios have high diversity while the human demonstration data is limited.