Ergodic Imitation for Adaptive Exploration around Demonstrations

TL;DR

This paper introduces an adaptive ergodic imitation method for robotics that improves exploration by interpolating between demonstration tracking and adaptive exploration, addressing environmental mismatches.

Contribution

It extends ergodic control to incorporate demonstrations within a retrieval-based framework for adaptive, online exploration in imitation learning.

Findings

Constructs a target distribution from demonstrations for adaptive trajectory generation.

Interpolates between demonstration tracking and exploration based on environmental conditions.

Enhances robustness of robotic imitation learning under environmental mismatches.

Abstract

In robotics, a common challenge in imitation learning is the mismatch between training and deployment conditions, caused, for example, by environmental changes or imperfect observation and control. When a robot follows a nominal trajectory under such mismatch, it may become stuck and fail to complete the task. This calls for adaptive online exploration strategies that remain grounded in demonstrations. To this end, we propose an adaptive ergodic imitation approach that constructs a target distribution from the geometry of the retrieved demonstrations and uses it to generate trajectories that adaptively interpolate between tracking and exploration. Our method extends ergodic control beyond its traditional role in area-coverage and search by incorporating demonstrations into a retrieval-based receding-horizon framework for adaptive imitation.