World Model-Based End-to-End Scene Generation for Accident Anticipation in Autonomous Driving

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Guan, Yanchen, Liao, Haicheng, Wang, Chengyue, Liu, Xingcheng, Zhang, Jiaxun, Li, Zhenning
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909692507193344
author Guan, Yanchen
Liao, Haicheng
Wang, Chengyue
Liu, Xingcheng
Zhang, Jiaxun
Li, Zhenning
author_facet Guan, Yanchen
Liao, Haicheng
Wang, Chengyue
Liu, Xingcheng
Zhang, Jiaxun
Li, Zhenning
contents Reliable anticipation of traffic accidents is essential for advancing autonomous driving systems. However, this objective is limited by two fundamental challenges: the scarcity of diverse, high-quality training data and the frequent absence of crucial object-level cues due to environmental disruptions or sensor deficiencies. To tackle these issues, we propose a comprehensive framework combining generative scene augmentation with adaptive temporal reasoning. Specifically, we develop a video generation pipeline that utilizes a world model guided by domain-informed prompts to create high-resolution, statistically consistent driving scenarios, particularly enriching the coverage of edge cases and complex interactions. In parallel, we construct a dynamic prediction model that encodes spatio-temporal relationships through strengthened graph convolutions and dilated temporal operators, effectively addressing data incompleteness and transient visual noise. Furthermore, we release a new benchmark dataset designed to better capture diverse real-world driving risks. Extensive experiments on public and newly released datasets confirm that our framework enhances both the accuracy and lead time of accident anticipation, offering a robust solution to current data and modeling limitations in safety-critical autonomous driving applications.
format Preprint
id arxiv_https___arxiv_org_abs_2507_12762
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle World Model-Based End-to-End Scene Generation for Accident Anticipation in Autonomous Driving
Guan, Yanchen
Liao, Haicheng
Wang, Chengyue
Liu, Xingcheng
Zhang, Jiaxun
Li, Zhenning
Computer Vision and Pattern Recognition
Machine Learning
Reliable anticipation of traffic accidents is essential for advancing autonomous driving systems. However, this objective is limited by two fundamental challenges: the scarcity of diverse, high-quality training data and the frequent absence of crucial object-level cues due to environmental disruptions or sensor deficiencies. To tackle these issues, we propose a comprehensive framework combining generative scene augmentation with adaptive temporal reasoning. Specifically, we develop a video generation pipeline that utilizes a world model guided by domain-informed prompts to create high-resolution, statistically consistent driving scenarios, particularly enriching the coverage of edge cases and complex interactions. In parallel, we construct a dynamic prediction model that encodes spatio-temporal relationships through strengthened graph convolutions and dilated temporal operators, effectively addressing data incompleteness and transient visual noise. Furthermore, we release a new benchmark dataset designed to better capture diverse real-world driving risks. Extensive experiments on public and newly released datasets confirm that our framework enhances both the accuracy and lead time of accident anticipation, offering a robust solution to current data and modeling limitations in safety-critical autonomous driving applications.
title World Model-Based End-to-End Scene Generation for Accident Anticipation in Autonomous Driving
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2507.12762