摘要
脑电图(Electroencephalography,EEG)信号分类在情感识别和脑机接口(Brain-computer interface,BCI)应用中具有关键意义。提出了一种参数共享的多特征图内外交互模型(Cross-map token attention,CMTA)。采用时空特征卷积神经网络(Spatial-temporal convolutional neural network,STCNN)对脑电图进行处理,生成多个脑电图特征图,每张特征图被视为一个token,传入参数共享的多模态模块MT(MLP和Transformer),其中多层感知器(Multi-layer perceptron,MLP)用于捕捉特征图内部的交互关系,Transformer则实现特征图之间的信息交互,从而提取更丰富的特征。通过一维自适应池化和全连接层构成的自适应分类器(Adapt-Classifier)完成脑电图的分类。实验结果表明,该方法在情感识别SEED数据集上的分类精度为98.86%,Kappa值为0.9829;在运动分类BCI Competition IV Dataset 2a数据集上的分类精度为81.20%,Kappa值为0.7484;在运动分类BCI Competition IV Dataset 2b数据集上的分类精度为86.55%,Kappa值为0.7352。实验结果验证了所提方法在脑电图分类任务中的优越性能,并展示了其在不同EEG数据集上的广泛适用性。
Electroencephalography(EEG)signal classification plays a crucial role in emotion recognition and brain-computer interface(BCI)applications.This paper proposes a parameter-sharing cross-map token attention(CMTA)model for intra-and inter-feature map interaction.Firstly,a spatial-temporal convolutional neural network(STCNN)is used to process EEG data,generating multiple EEG feature maps.Each feature map is treated as a token and fed into a parameter-sharing multi-modal module MT,which integrates a multi-layer perceptron(MLP)and a Transformer.The MLP captures intra-feature map interactions,while the Transformer enables information exchange between feature maps,thereby extracting richer features.Finally,an adaptive classifier(Adapt-Classifier)consisting of one-dimensional adaptive pooling and a fully connected layer is used to perform EEG classification.Experimental results show that the proposed method achieves a classification accuracy of 98.86%and a Kappa value of 0.9829 on the SEED dataset for emotion recognition,an accuracy of 81.20%and a Kappa value of 0.7484 on the BCI Competition IV Dataset 2a for motor imagery classification,and an accuracy of 86.55%and a Kappa value of 0.7352 on the BCI Competition IV Dataset 2b.These results demonstrate the superior performance of the proposed method in EEG classification tasks and highlight its broad applicability across different EEG datasets.
作者
闭应洲
刘善锐
霍雷刚
甘秋静
李永玉
BI Yingzhou;LIU Shanrui;HUO Leigang;GAN Qiujing;LI Yongyu(School of Artificial Intelligence,Nanning Normal University,Nanning 530199,China)
出处
《数据采集与处理》
北大核心
2025年第4期950-961,共12页
Journal of Data Acquisition and Processing
基金
国家自然科学基金(62067007)
广西学位与研究生教改课题(JGY2023236)。
