ITSC 2025 Paper Abstract

Paper WE-EA-T14.2

Zhang, Wenqi (Tsinghua University), Qin, Yanjun (Xinjiang University), Tao, Xiaoming (Tsinghua University)

Driver Vigilance Detection Based on Single-Channel EEG with a Multi-Scale Convolution Transformer

Scheduled for presentation during the Regular Session "S14b-Human Factors and Human Machine Interaction in Automated Driving" (WE-EA-T14), Wednesday, November 19, 2025, 13:50−14:10, Currumbin

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

Keywords Human-Machine Interaction Systems for Enhanced Driver Assistance and Safety, Driver Behavior Monitoring and Feedback Systems for Semi-autonomous Vehicles

Abstract

Driver vigilance is closely related to driving safety and can be monitored through human data. Research based on facial features is susceptible to influences such as light and obstructions, hence many studies have shifted towards physiological signals. However, electroencephalography (EEG)-based work faces challenges related to portability. Therefore, this paper is dedicated to studying drivers' vigilance based on single-channel EEG. Addressing this issue, we propose a multi-scale convolution transformer (MSCTSF) model for feature extraction and detection of single-channel EEG data to determine the driver's vigilance. We employ leave-one-subject-out cross-validation (LOSO-CV) to validate the model's cross-subject detection performance. Experimental results indicate that our model achieves a higher accuracy in detecting driver vigilance, reaching an average accuracy of 76.40%, surpassing existing advanced baselines. This research can advance the study of human factors in intelligent transportation systems and contribute to ensuring driving safety.