DOI码：10.1016/j.bspc.2026.110176

所属单位：南京航空航天大学

发表刊物：Biomedical Signal Processing and Control

关键字：Driver fatigue detection; Multilevel data fusion; Dempster-Shafer evidence theory; EEG signal; EOG signal; Deep neural networks

摘要：To enhance traffic safety, effective monitoring of driver fatigue is essential. However, unimodal methods can be unstable under noise, and many multimodal approaches still lack explicit uncertainty handling for safety – critical use. We propose a multilevel EEG-EOG fusion framework to improve fatigue-state recognition. Specifically, we extract Differential Entropy (DE) features from EEG signals to construct 4D representations encapsulating temporal, spectral, and spatial information, alongside 36 complementary statistical eye movement features derived from EOG signals. A CBAM-4D-CRNN

备注：文章号110176，发表日期待核实

论文类型：期刊论文

学科门类：工学

一级学科：航空宇航科学与技术

文献类型：J

卷号：120

页面范围：110176

ISSN号：1746-8094

是否译文：否

发表时间：2026-01-01

合写作者：Bowen Li,方美华,魏志勇,张铭

第一作者：丁牧恒,丁牧恒

通讯作者：丁牧恒,丁牧恒,丁牧恒