基于时空注意力机制的轻量级脑纹识别算法

方芳,严军,郭红想,王勇

Lightweight brainprint recognition algorithm based on spatio-temporal attention mechanism

Fang FANG,Jun YAN,Hongxiang GUO,Yong WANG

表 1 不同参数对2个数据集的不同状态数据的分类性能影响

Tab.1 Effect of different parameter on classification performance for different state data of two datasets %

实验参数 参数选择 ACC±std Physionet-MI Physionet-EO Physionet-EC DEAP 卷积核数 (16,64) (64,128) 99.43±0.15 99.97±0.09 98.13±0.17 99.45±0.15 97.32±0.41 99.01±0.30 100.00±0.02 100.00±0.00 激活函数 ELU SiLU GeLU ReLU 99.95±0.11 99.96±0.10 99.97±0.10 99.97±0.09 98.64±0.23 98.70±0.21 99.15±0.17 99.45±0.15 98.26±0.39 98.34±0.36 97.35±0.45 99.01±0.30 99.92±0.08 100.00±0.01 100.00±0.00 100.00±0.00 注意力 GCT CA 99.95±0.13 99.97±0.09 99.12±0.16 99.45±0.15 98.34±0.33 99.01±0.30 100.00±0.01 100.00±0.00 池化层 AvgPool2d MaxPool2d 99.97±0.10 99.97±0.09 98.94±0.18 99.45±0.15 98.59±0.33 99.01±0.30 100.00±0.00 100.00±0.00 Dropout 0.25 0.5 99.91±0.12 99.97±0.09 99.06±0.16 99.45±0.15 98.77±0.32 99.01±0.30 100.00±0.01 100.00±0.00