SocialDriveGen: Generating Diverse Traffic Scenarios with Controllable Social Interactions

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Tian, Jiaguo, Zhu, Zhengbang, Zhang, Shenyu, Xu, Li, Zheng, Bo, Liu, Xu, Peng, Weiji, Yao, Shizeng, Zhang, Weinan
Format:	Preprint
Published:	2025
Subjects:	Multiagent Systems Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866915646917312512
author	Tian, Jiaguo Zhu, Zhengbang Zhang, Shenyu Xu, Li Zheng, Bo Liu, Xu Peng, Weiji Yao, Shizeng Zhang, Weinan
author_facet	Tian, Jiaguo Zhu, Zhengbang Zhang, Shenyu Xu, Li Zheng, Bo Liu, Xu Peng, Weiji Yao, Shizeng Zhang, Weinan
contents	The generation of realistic and diverse traffic scenarios in simulation is essential for developing and evaluating autonomous driving systems. However, most simulation frameworks rely on rule-based or simplified models for scene generation, which lack the fidelity and diversity needed to represent real-world driving. While recent advances in generative modeling produce more realistic and context-aware traffic interactions, they often overlook how social preferences influence driving behavior. SocialDriveGen addresses this gap through a hierarchical framework that integrates semantic reasoning and social preference modeling with generative trajectory synthesis. By modeling egoism and altruism as complementary social dimensions, our framework enables controllable diversity in driver personalities and interaction styles. Experiments on the Argoverse 2 dataset show that SocialDriveGen generates diverse, high-fidelity traffic scenarios spanning cooperative to adversarial behaviors, significantly enhancing policy robustness and generalization to rare or high-risk situations.
format	Preprint
id	arxiv_https___arxiv_org_abs_2512_01363
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	SocialDriveGen: Generating Diverse Traffic Scenarios with Controllable Social Interactions Tian, Jiaguo Zhu, Zhengbang Zhang, Shenyu Xu, Li Zheng, Bo Liu, Xu Peng, Weiji Yao, Shizeng Zhang, Weinan Multiagent Systems Machine Learning The generation of realistic and diverse traffic scenarios in simulation is essential for developing and evaluating autonomous driving systems. However, most simulation frameworks rely on rule-based or simplified models for scene generation, which lack the fidelity and diversity needed to represent real-world driving. While recent advances in generative modeling produce more realistic and context-aware traffic interactions, they often overlook how social preferences influence driving behavior. SocialDriveGen addresses this gap through a hierarchical framework that integrates semantic reasoning and social preference modeling with generative trajectory synthesis. By modeling egoism and altruism as complementary social dimensions, our framework enables controllable diversity in driver personalities and interaction styles. Experiments on the Argoverse 2 dataset show that SocialDriveGen generates diverse, high-fidelity traffic scenarios spanning cooperative to adversarial behaviors, significantly enhancing policy robustness and generalization to rare or high-risk situations.
title	SocialDriveGen: Generating Diverse Traffic Scenarios with Controllable Social Interactions
topic	Multiagent Systems Machine Learning
url	https://arxiv.org/abs/2512.01363