ReSim: Reliable World Simulation for Autonomous Driving

TL;DR

ReSim is a controllable, high-fidelity world simulation model for autonomous driving that incorporates diverse behaviors, including hazardous non-expert actions, to improve policy evaluation and planning.

Contribution

It introduces a novel diffusion transformer-based world model trained on heterogeneous data, including simulator-generated non-expert behaviors, with a Video2Reward module for reward estimation.

Findings

ReSim achieves up to 44% higher visual fidelity.

Controllability improves over 50% for various actions.

Planning performance on NAVSIM increases by 25%.

Abstract

How can we reliably simulate future driving scenarios under a wide range of ego driving behaviors? Recent driving world models, developed exclusively on real-world driving data composed mainly of safe expert trajectories, struggle to follow hazardous or non-expert behaviors, which are rare in such data. This limitation restricts their applicability to tasks such as policy evaluation. In this work, we address this challenge by enriching real-world human demonstrations with diverse non-expert data collected from a driving simulator (e.g., CARLA), and building a controllable world model trained on this heterogeneous corpus. Starting with a video generator featuring a diffusion transformer architecture, we devise several strategies to effectively integrate conditioning signals and improve prediction controllability and fidelity. The resulting model, ReSim, enables Reliable Simulation of diverse open-world driving scenarios under various actions, including hazardous non-expert ones. To close the gap between high-fidelity simulation and applications that require reward signals to judge different actions, we introduce a Video2Reward module that estimates a reward from ReSim's simulated future. Our ReSim paradigm achieves up to 44% higher visual fidelity, improves controllability for both expert and non-expert actions by over 50%, and boosts planning and policy selection performance on NAVSIM by 2% and 25%, respectively.