Path Diffuser: Diffusion Model for Data-Driven Traffic Simulator

2025-09-29

AI Summary

To address insufficient trajectory diversity and poor road compliance in out-of-distribution (OOD) maps—both stemming from overreliance on historical agent trajectories—this paper proposes a two-stage diffusion model. In the first stage, candidate trajectories are generated in the Frenet coordinate system under explicit road constraints, leveraging prior knowledge of map geometry and motion primitives. The second stage models multi-agent interactions to jointly initialize poses and generate trajectories without contextual conditioning. Crucially, the method operates solely on static map inputs, eliminating dependence on historical agent data. Experiments on Argoverse 2 demonstrate significant improvements: a 1.92× gain in trajectory distribution quality, a 1.14× improvement in commonsense plausibility, and a 1.62× increase in road compliance. These advances substantially enhance the generalization capability of autonomous driving planners during zero-shot or OOD evaluation.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Multiagent Systems: Multiagent Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Simulating diverse and realistic traffic scenarios is critical for developing and testing autonomous planning. Traditional rule-based planners lack diversity and realism, while learning-based simulators often replay, forecast, or edit scenarios using historical agent trajectories. However, they struggle to generate new scenarios, limiting scalability and diversity due to their reliance on fully annotated logs and historical data. Thus, a key challenge for a learning-based simulator's performance is that it requires agents' past trajectories and pose information in addition to map data, which might not be available for all agents on the road.Without which, generated scenarios often produce unrealistic trajectories that deviate from drivable areas, particularly under out-of-distribution (OOD) map scenes (e.g., curved roads). To address this, we propose Path Diffuser (PD): a two-stage, diffusion model for generating agent pose initializations and their corresponding trajectories conditioned on the map, free of any historical context of agents' trajectories. Furthermore, PD incorporates a motion primitive-based prior, leveraging Frenet frame candidate trajectories to enhance diversity while ensuring road-compliant trajectory generation. We also explore various design choices for modeling complex multi-agent interactions. We demonstrate the effectiveness of our method through extensive experiments on the Argoverse2 Dataset and additionally evaluate the generalizability of the approach on OOD map variants. Notably, Path Diffuser outperforms the baseline methods by 1.92x on distribution metrics, 1.14x on common-sense metrics, and 1.62x on road compliance from adversarial benchmarks.

Problem

• Generating realistic traffic scenarios without historical trajectory data
• Overcoming limitations of rule-based and learning-based traffic simulators
• Ensuring road-compliant trajectory generation under diverse map conditions

Innovation

• Two-stage diffusion model generates trajectories without historical data
• Motion primitive prior ensures road-compliant trajectory generation
• Handles multi-agent interactions and out-of-distribution maps