基于通道加权的多模态特征融合用于EEG疲劳驾驶检测
程文鑫,闫光辉,常文文,吴佰靖,黄亚宁

Channel-weighted multimodal feature fusion for EEG-based fatigue driving detection
Wenxin CHENG,Guanghui YAN,Wenwen CHANG,Baijing WU,Yaning HUANG
表 5 相同数据集及相同分类任务下不同模型的对比实验结果
Tab.5 Comparative experimental results of different models with the same dataset and the same classification task
方法APSMF1
D1D2D1D2D1D2D1D2D1D2
RF[26]0.740 00.668 00.740 00.698 90.540 00.639 80.260 00.332 00.540 00.611 1
LSTM[27]0.881 30.853 50.914 70.889 00.884 60.877 40.118 70.146 50.898 20.857 7
EEG-Conv[28]0.779 60.802 20.853 90.882 60.585 30.662 50.220 40.197 80.632 70.652 4
EEG-TCNet[29]0.813 30.851 50.697 70.800 50.674 30.700 40.186 70.148 50.674 30.708 8
EEGNet[30]0.798 90.643 50.745 20.625 50.752 20.653 10.201 10.356 50.744 20.621 1
ESTCNN[31]0.790 60.772 40.767 10.797 90.789 80.766 50.209 40.227 60.779 60.795 8
CSF-GTNet[15]0.821 10.841 50.772 30.833 40.843 30.841 90.178 90.158 50.780 90.814 9
T-A-MFFNet[5]0.858 60.865 40.733 90.892 30.826 50.843 70.141 40.134 60.777 40.800 5
本研究方法0.933 70.907 80.955 90.924 60.934 50.919 70.066 30.092 20.945 10.922 1