Bayesian Imitation Learning for End-to-End Mobile Manipulation

TL;DR

This paper presents a Bayesian imitation learning approach that combines multiple sensor inputs to improve robot manipulation tasks, effectively reducing the sim-to-real gap and enhancing generalization in real-world environments.

Contribution

It introduces a variational information bottleneck regularization method for multi-sensor imitation learning, improving domain transfer and sensor fusion in robotic manipulation.

Findings

Achieved 96% task success rate in real-world office environment.

Reduced sim-to-real gap through sensor-agnostic regularization.

Enhanced sensor fusion by understanding situational uncertainty.

Abstract

In this work we investigate and demonstrate benefits of a Bayesian approach to imitation learning from multiple sensor inputs, as applied to the task of opening office doors with a mobile manipulator. Augmenting policies with additional sensor inputs, such as RGB + depth cameras, is a straightforward approach to improving robot perception capabilities, especially for tasks that may favor different sensors in different situations. As we scale multi-sensor robotic learning to unstructured real-world settings (e.g. offices, homes) and more complex robot behaviors, we also increase reliance on simulators for cost, efficiency, and safety. Consequently, the sim-to-real gap across multiple sensor modalities also increases, making simulated validation more difficult. We show that using the Variational Information Bottleneck (Alemi et al., 2016) to regularize convolutional neural networks improves generalization to held-out domains and reduces the sim-to-real gap in a sensor-agnostic manner. As a side effect, the learned embeddings also provide useful estimates of model uncertainty for each sensor. We demonstrate that our method is able to help close the sim-to-real gap and successfully fuse RGB and depth modalities based on understanding of the situational uncertainty of each sensor. In a real-world office environment, we achieve 96% task success, improving upon the baseline by +16%.