Driving Attention State Detection Based on GRU-EEGNet

• DOI: 10.3390/s24165086
• Publication Date: 2024-08-07T12:42:28Z
• Authors: Xiaoli Wu; Changcheng Shi; Lirong Yan
• Keywords: Driving simulator; Distracted driving

The present study utilizes the significant differences in θ, α, and β band power spectra observed electroencephalograms (EEGs) during distracted versus focused driving. Three subtasks, visual distraction, auditory cognitive were designed to appear randomly driving simulations. of EEG signals four attention states extracted, SVM, EEGNet, GRU-EEGNet models employed for detection states, respectively. Online experiments conducted. extraction spectrum features was found be a more effective method than whole states. state accuracy proposed model is improved by 6.3% 12.8% over EEGNet PSD_SVM method, decoding combining an deep learning algorithm, which effectively improves accuracy, manually preliminarily selected based on results existing studies.