Scenario Diffusion: Controllable Driving Scenario Generation With Diffusion
This addresses the need for automated scenario generation in autonomous vehicle testing, though it appears incremental as it adapts existing diffusion methods to this domain.
The paper tackles the problem of generating synthetic traffic scenarios for autonomous vehicle safety validation by proposing Scenario Diffusion, a diffusion-based architecture that generates controllable distributions of agent poses, orientations, and trajectories conditioned on maps and scenario tokens, achieving generalization across different geographical regions.
Automated creation of synthetic traffic scenarios is a key part of validating the safety of autonomous vehicles (AVs). In this paper, we propose Scenario Diffusion, a novel diffusion-based architecture for generating traffic scenarios that enables controllable scenario generation. We combine latent diffusion, object detection and trajectory regression to generate distributions of synthetic agent poses, orientations and trajectories simultaneously. To provide additional control over the generated scenario, this distribution is conditioned on a map and sets of tokens describing the desired scenario. We show that our approach has sufficient expressive capacity to model diverse traffic patterns and generalizes to different geographical regions.