Leng, Wenfeng (Shanghai Jiao Tong University), Wang, Yakang (Shanghai Jiao Tong University), Huang, Yonghui (Shanghai Jiao Tong University), Hu, Chuan (Shanghai Jiao Tong University), Zhang, Xi (Shanghai Jiao Tong University)

Overcoming Communication Time Delay in V2V Collaborative Perception

Scheduled for presentation during the Regular Session "S10b-Cooperative and Connected Autonomous Systems" (WE-EA-T10), Wednesday, November 19, 2025, 13:50−14:10, Cooleangata 4

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 October 19, 2025

Abstract

Collaborative perception can achieve super-visual perception and solve the problems of blind zones and detection distance limitation in single-vehicle perception. However, in actual application scenarios, the limited communication bandwidth results in time delay in data transmission between vehicles. The features with time delay received by the ego vehicle cannot fully exert the potential of collaborative perception. To effectively utilize the features with time delay from other vehicles, we propose a two order flow generator based on a physical motion model, which can accurately predict the current features using historical data, thereby improving the performance of collaborative perception. In addition, we specially design a two stage training strategy and a loss function for the entire model. Experimental results show that our proposed method outperforms the current Collaborative perception methods on the V2XSet dataset and can effectively overcome the problem of time delay.