Hierarchical Model-Based Imitation Learning for Planning in Autonomous Driving

TL;DR

This paper introduces a hierarchical model-based imitation learning approach for autonomous driving, enabling robust zero-shot generalization to new routes in dense urban environments using large-scale real-world data.

Contribution

The work extends MGAIL with a hierarchical model, allowing generalization to arbitrary goals and demonstrating zero-shot transfer in complex urban driving scenarios.

Findings

Achieved robust zero-shot navigation in synthetic scenarios

Mixing MGAIL with behavior cloning improves policy performance

Policy approaches expert-level driving performance

Abstract

We demonstrate the first large-scale application of model-based generative adversarial imitation learning (MGAIL) to the task of dense urban self-driving. We augment standard MGAIL using a hierarchical model to enable generalization to arbitrary goal routes, and measure performance using a closed-loop evaluation framework with simulated interactive agents. We train policies from expert trajectories collected from real vehicles driving over 100,000 miles in San Francisco, and demonstrate a steerable policy that can navigate robustly even in a zero-shot setting, generalizing to synthetic scenarios with novel goals that never occurred in real-world driving. We also demonstrate the importance of mixing closed-loop MGAIL losses with open-loop behavior cloning losses, and show our best policy approaches the performance of the expert. We evaluate our imitative model in both average and challenging scenarios, and show how it can serve as a useful prior to plan successful trajectories.