Search papers, labs, and topics across Lattice.
This paper introduces a novel diffusion-based framework for generating closed-loop traffic scenarios that are both realistic and controllable, addressing the challenges of interactive multi-agent behaviors in autonomous vehicle simulations. By leveraging instance-centric scene context and multimodal proposal priors, the method enhances sampling efficiency and reduces computational costs, making it suitable for real-time applications. Experiments on the Waymo Open Motion Dataset reveal that the approach achieves a desirable balance between realism, safety, and controllability, with test-time guidance allowing for strategic trade-offs among competing objectives.
Achieving real-time, controllable traffic scenario generation without sacrificing realism could revolutionize autonomous vehicle simulations.
Closed-loop traffic simulation remains challenging because it must generate interactive multi-agent behaviors that are scene-consistent and controllable throughout rollout. Prior diffusion-based approaches achieve strong realism, but their computational cost can hinder deployment in time-constrained replanning loops for autonomous vehicle planning and simulation. We present a diffusion-based scenario generation framework conditioned on instance-centric scene context and multimodal proposal priors, with optional test-time guidance for shaping safety-critical behaviors. A compact action-latent representation and proposal-based initialization improve sampling efficiency and reduce per-step runtime without retraining. Experiments on the Waymo Open Motion Dataset demonstrate a favorable balance among realism, safety, and controllability across diverse interactive scenarios, while showing that test-time guidance enables systematic trade-offs among competing objectives.
