MRIC: Model-Based Reinforcement-Imitation Learning with Mixture-of-Codebooks for Autonomous Driving Simulation

TL;DR

MRIC is a novel model-based reinforcement-imitation learning framework that uses mixture-of-codebooks and regularizations to simulate diverse, realistic behaviors of heterogeneous agents in autonomous driving scenarios, improving over state-of-the-art methods.

Contribution

The paper introduces MRIC, a new framework combining mixture-of-codebooks with dual regularizations for stable, diverse, and realistic autonomous driving simulation, addressing multi-modality and high-dimensionality challenges.

Findings

MRIC outperforms baselines on diversity and realism metrics.

MRIC achieves lower collision rates and better distributional similarity.

The approach effectively models heterogeneous agent behaviors in complex scenarios.

Abstract

Accurately simulating diverse behaviors of heterogeneous agents in various scenarios is fundamental to autonomous driving simulation. This task is challenging due to the multi-modality of behavior distribution, the high-dimensionality of driving scenarios, distribution shift, and incomplete information. Our first insight is to leverage state-matching through differentiable simulation to provide meaningful learning signals and achieve efficient credit assignment for the policy. This is demonstrated by revealing the existence of gradient highways and interagent gradient pathways. However, the issues of gradient explosion and weak supervision in low-density regions are discovered. Our second insight is that these issues can be addressed by applying dual policy regularizations to narrow the function space. Further considering diversity, our third insight is that the behaviors of heterogeneous agents in the dataset can be effectively compressed as a series of prototype vectors for retrieval. These lead to our model-based reinforcement-imitation learning framework with temporally abstracted mixture-of-codebooks (MRIC). MRIC introduces the open-loop modelbased imitation learning regularization to stabilize training, and modelbased reinforcement learning (RL) regularization to inject domain knowledge. The RL regularization involves differentiable Minkowskidifference-based collision avoidance and projection-based on-road and traffic rule compliance rewards. A dynamic multiplier mechanism is further proposed to eliminate the interference from the regularizations while ensuring their effectiveness. Experimental results using the largescale Waymo open motion dataset show that MRIC outperforms state-ofthe-art baselines on diversity, behavioral realism, and distributional realism, with large margins on some key metrics (e.g., collision rate, minSADE, and time-to-collision JSD).