Prior knowledge-guided multi-information graph convolutional network for driver drowsiness detection

Open Webpage

Feng Wei, Jucheng Yang, Yuan Wang, Liang Lin, Haibin Zhang

Published: 2025, Last Modified: 05 Nov 2025Expert Syst. Appl. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•MIGCN leverages prior knowledge to address driver variability, outperforming current methods.•MIGCN uses multi-source features to reduce single-feature sensitivity and improve accuracy.•MIGCN builds a graph with multi-information features, helping GCN learn spatial relations.