Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Rowe, Luke, Girgis, Roger, Gosselin, Anthony, Paull, Liam, Pal, Christopher, Heide, Felix
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912299495718912
author Rowe, Luke
Girgis, Roger
Gosselin, Anthony
Paull, Liam
Pal, Christopher
Heide, Felix
author_facet Rowe, Luke
Girgis, Roger
Gosselin, Anthony
Paull, Liam
Pal, Christopher
Heide, Felix
contents We introduce Scenario Dreamer, a fully data-driven generative simulator for autonomous vehicle planning that generates both the initial traffic scene - comprising a lane graph and agent bounding boxes - and closed-loop agent behaviours. Existing methods for generating driving simulation environments encode the initial traffic scene as a rasterized image and, as such, require parameter-heavy networks that perform unnecessary computation due to many empty pixels in the rasterized scene. Moreover, we find that existing methods that employ rule-based agent behaviours lack diversity and realism. Scenario Dreamer instead employs a novel vectorized latent diffusion model for initial scene generation that directly operates on the vectorized scene elements and an autoregressive Transformer for data-driven agent behaviour simulation. Scenario Dreamer additionally supports scene extrapolation via diffusion inpainting, enabling the generation of unbounded simulation environments. Extensive experiments show that Scenario Dreamer outperforms existing generative simulators in realism and efficiency: the vectorized scene-generation base model achieves superior generation quality with around 2x fewer parameters, 6x lower generation latency, and 10x fewer GPU training hours compared to the strongest baseline. We confirm its practical utility by showing that reinforcement learning planning agents are more challenged in Scenario Dreamer environments than traditional non-generative simulation environments, especially on long and adversarial driving environments.
format Preprint
id arxiv_https___arxiv_org_abs_2503_22496
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments
Rowe, Luke
Girgis, Roger
Gosselin, Anthony
Paull, Liam
Pal, Christopher
Heide, Felix
Robotics
Computer Vision and Pattern Recognition
We introduce Scenario Dreamer, a fully data-driven generative simulator for autonomous vehicle planning that generates both the initial traffic scene - comprising a lane graph and agent bounding boxes - and closed-loop agent behaviours. Existing methods for generating driving simulation environments encode the initial traffic scene as a rasterized image and, as such, require parameter-heavy networks that perform unnecessary computation due to many empty pixels in the rasterized scene. Moreover, we find that existing methods that employ rule-based agent behaviours lack diversity and realism. Scenario Dreamer instead employs a novel vectorized latent diffusion model for initial scene generation that directly operates on the vectorized scene elements and an autoregressive Transformer for data-driven agent behaviour simulation. Scenario Dreamer additionally supports scene extrapolation via diffusion inpainting, enabling the generation of unbounded simulation environments. Extensive experiments show that Scenario Dreamer outperforms existing generative simulators in realism and efficiency: the vectorized scene-generation base model achieves superior generation quality with around 2x fewer parameters, 6x lower generation latency, and 10x fewer GPU training hours compared to the strongest baseline. We confirm its practical utility by showing that reinforcement learning planning agents are more challenged in Scenario Dreamer environments than traditional non-generative simulation environments, especially on long and adversarial driving environments.
title Scenario Dreamer: Vectorized Latent Diffusion for Generating Driving Simulation Environments
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.22496