sis 18: e8

Research Article

Channel space weighted fusion-oriented feature pyramid network for motor imagery EEG signal recognition

Download36 downloads
  • @ARTICLE{10.4108/eai.3-9-2021.170906,
        author={Wenhao Yang},
        title={Channel space weighted fusion-oriented feature pyramid network for motor imagery EEG signal recognition},
        journal={EAI Endorsed Transactions on Scalable Information Systems: Online First},
        volume={},
        number={},
        publisher={EAI},
        journal_a={SIS},
        year={2021},
        month={9},
        keywords={motor imagery EEG signal recognition, short-time Fourier transform, attention mechanism, FPN, channel spatial weighted fusion},
        doi={10.4108/eai.3-9-2021.170906}
    }
    
  • Wenhao Yang
    Year: 2021
    Channel space weighted fusion-oriented feature pyramid network for motor imagery EEG signal recognition
    SIS
    EAI
    DOI: 10.4108/eai.3-9-2021.170906
Wenhao Yang1,*
  • 1: School of Physical Education and Sport, Zhengzhou University of Science and Technology, Zhengzhou, 450064, China
*Contact email: zxcvfdsa5024@foxmail.com

Abstract

In order to solve the problems of weak generalization ability and low classification accuracy in motor imagery EEG signal classification, this paper proposes a channel space weighted fusion-oriented feature pyramid network for motor imagery EEG signal recognition. First, the short-time Fourier transform is used to obtain the EEG time-frequency map. Then, it builds a new feature pyramid network(FPN). The attention mechanism module is integrated into the FPN module, and the channel spatial weighted fusion-oriented feature pyramid network is proposed. This new structure can not only learn the weight of important channel features in the feature map, but also learn the representation of important feature areas in the network layers. Meanwhile, Skip-FPN module is added into the network structure, which fuses more details of EEG signals through short connections. The Dropout layer is added to prevent network training from over-fitting. In the classification model, we improve the AdaBoost algorithm to automatically update the base learner according to the classification error rate. Finally, the proposed model is used to classify the test data and the Kappa value is used as the evaluation index. Compared with the state-of-the-art motor image EEG signal recognition methods, the proposed method achieves better performance on the BCI Competition IV 2b data set. It has good generalization ability and can improve the classification effect.