Zero-shot Imitation Learning from Demonstrations for Legged Robot Visual Navigation

TL;DR

This paper introduces a zero-shot imitation learning framework for legged robot navigation from third-person human demonstrations, addressing perspective differences and dynamic variations, enabling effective real-world navigation without path-specific training.

Contribution

The paper presents a novel zero-shot imitation learning method that uses feature disentanglement and inverse modeling to train legged robot navigation policies from third-person demonstrations.

Findings

Effective navigation policy learned for Laikago robot

Framework works in both simulated and real environments

Zero-shot learning without path-specific training

Abstract

Imitation learning is a popular approach for training visual navigation policies. However, collecting expert demonstrations for legged robots is challenging as these robots can be hard to control, move slowly, and cannot operate continuously for a long time. Here, we propose a zero-shot imitation learning approach for training a visual navigation policy on legged robots from human (third-person perspective) demonstrations, enabling high-quality navigation and cost-effective data collection. However, imitation learning from third-person demonstrations raises unique challenges. First, these demonstrations are captured from different camera perspectives, which we address via a feature disentanglement network (FDN) that extracts perspective-invariant state features. Second, as transition dynamics vary across systems, we label missing actions by either building an inverse model of the robot's dynamics in the feature space and applying it to the human demonstrations or developing a Graphic User Interface(GUI) to label human demonstrations. To train a navigation policy we use a model-based imitation learning approach with FDN and labeled human demonstrations. We show that our framework can learn an effective policy for a legged robot, Laikago, from human demonstrations in both simulated and real-world environments. Our approach is zero-shot as the robot never navigates the same paths during training as those at testing time. We justify our framework by performing a comparative study.