Luo, Zhiyao (Institute of Automation, Chinese Academy of Sciences), Tian, Yonglin (Institute of Automation, Chinese Academy of Sciences), Bai, Songlin (University of Science and Technology Beijing), Wu, Xianke (Beijing University of Posts and Telecommunications), Liu, Yu-Liang (Huawei), Lv, Yisheng (Institute of Automation, Chinese Academy of Sciences), Wang, Fei-Yue (Institute of Automation, Chinese Academy of Sciences)

VIRTUE: Visual Interference from Roadside Billboard Targets on Urban End-To-End Autonomous Driving

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

Vision-based autonomous vehicles (AVs) rely heavily on precise perception of surrounding objects like vehicles and pedestrians which plays a crucial role on AVs' planning decisions. However, recent news reports have suggested that vehicles or pedestrians depicted on roadside billboards can potentially deceive vision-based autonomous driving systems, causing misclassification as real vehicles driving on the road and triggering abrupt steering or emergency breaks, which poses significant traffic safety risks. To quantitatively assess the impact of such billboard patterns on vision-based AV planning, we conducted evaluations on several existing end-to-end planning methods using synthetic image data where the positions, poses and patterns of the billboards are manually determined by us. Our evaluation demonstrates that billboards and the vehicles depicted on billboards indeed influence AV planning, with certain specific positions of billboards and certain specific angles of depicted vehicles exhibiting greater impacts.