Benchmarking Augmentation Methods for Learning Robust Navigation Agents: the Winning Entry of the 2021 iGibson Challenge

TL;DR

This paper benchmarks augmentation techniques for training robust navigation agents in dynamic environments, demonstrating that adding moving obstacles during training significantly improves generalization and success rates, especially in challenging scenarios.

Contribution

It introduces and evaluates dynamic obstacle augmentation methods, showing their effectiveness in enhancing navigation agent robustness and success in the 2021 iGibson Challenge.

Findings

Dynamic obstacle augmentation improves test-time generalization.

Combining image and obstacle augmentation yields higher success rates.

The approach is more robust to sim-to-sim transfer.

Abstract

Recent advances in deep reinforcement learning and scalable photorealistic simulation have led to increasingly mature embodied AI for various visual tasks, including navigation. However, while impressive progress has been made for teaching embodied agents to navigate static environments, much less progress has been made on more dynamic environments that may include moving pedestrians or movable obstacles. In this study, we aim to benchmark different augmentation techniques for improving the agent's performance in these challenging environments. We show that adding several dynamic obstacles into the scene during training confers significant improvements in test-time generalization, achieving much higher success rates than baseline agents. We find that this approach can also be combined with image augmentation methods to achieve even higher success rates. Additionally, we show that this approach is also more robust to sim-to-sim transfer than image augmentation methods. Finally, we demonstrate the effectiveness of this dynamic obstacle augmentation approach by using it to train an agent for the 2021 iGibson Challenge at CVPR, where it achieved 1st place for Interactive Navigation. Video link: https://www.youtube.com/watch?v=HxUX2HeOSE4