RADE: Learning Risk-Adjustable Driving Environment via Multi-Agent Conditional Diffusion

2025-05-06

AI Summary

Existing methods generate safety-critical scenarios via single-agent trajectory perturbation, compromising realism and scalability. Method: This paper proposes a multi-agent conditional diffusion model that employs agent-level risk metrics as learnable conditioning variables to jointly synthesize high-fidelity test scenarios conforming to real-world traffic statistics while enabling precise risk control. Contribution/Results: It eliminates handcrafted adversarial design, enabling data-driven natural interaction and collaborative risk modeling; introduces a motion tokenization-based kinematic validation module to ensure physical plausibility. Evaluated on the rounD dataset, the method preserves authentic statistical characteristics across risk levels, achieves monotonic increase in safety-critical event rates with prescribed risk levels, and demonstrates superior generation quality, controllability, and scalability.

Technology Category

• Multiagent Systems: Adversarial Agents
• Natural Language Processing: Safety and Robustness
• Computer Vision: Diffusion Models for Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Generating safety-critical scenarios in high-fidelity simulations offers a promising and cost-effective approach for efficient testing of autonomous vehicles. Existing methods typically rely on manipulating a single vehicle's trajectory through sophisticated designed objectives to induce adversarial interactions, often at the cost of realism and scalability. In this work, we propose the Risk-Adjustable Driving Environment (RADE), a simulation framework that generates statistically realistic and risk-adjustable traffic scenes. Built upon a multi-agent diffusion architecture, RADE jointly models the behavior of all agents in the environment and conditions their trajectories on a surrogate risk measure. Unlike traditional adversarial methods, RADE learns risk-conditioned behaviors directly from data, preserving naturalistic multi-agent interactions with controllable risk levels. To ensure physical plausibility, we incorporate a tokenized dynamics check module that efficiently filters generated trajectories using a motion vocabulary. We validate RADE on the real-world rounD dataset, demonstrating that it preserves statistical realism across varying risk levels and naturally increases the likelihood of safety-critical events as the desired risk level grows up. Our results highlight RADE's potential as a scalable and realistic tool for AV safety evaluation.

Problem

• Generating realistic risk-adjustable traffic scenes
• Preserving naturalistic multi-agent interactions with controllable risk
• Ensuring physical plausibility in generated trajectories

Innovation

• Multi-agent diffusion models generate realistic traffic scenes
• Risk-conditioned behaviors learned directly from data
• Tokenized dynamics check ensures physical plausibility