A semi-supervised speech imagery intention decoding method based on electroencephalogram collaborative clustering

By employing a semi-supervised method based on EEG co-clustering, the problems of model transparency and overfitting in speech imagery intention decoding are solved, improving the interpretability and accuracy of decoding. This method is suitable for non-invasive signal acquisition and expands the application population and scenarios.

CN119557677BActive Publication Date: 2025-10-24HANGZHOU DIANZI UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202411598571.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-11-11
Publication Date
2025-10-24
Estimated Expiration
2044-11-11

Smart Images

  • Figure CN119557677B_ABST
    Figure CN119557677B_ABST
Patent Text Reader

Abstract

The application discloses a kind of semi-supervised speech imagination intention decoding methods based on electroencephalogram collaborative clustering, comprising the following steps: step 1, the electroencephalogram data of speech imagination is collected;Step 2, the feature extraction of preprocessed electroencephalogram data is carried out;Step 3, the original data matrix is carried out feature clustering and sample clustering, so as to obtain the clustering indication matrix of feature, the clustering indication matrix of sample and coefficient matrix;Step 4, the optimization solution of the clustering indication matrix of feature, the clustering indication matrix of sample and coefficient matrix is completed under semi-supervised framework;Step 5, by visual means, the decoding result obtained based on electroencephalogram sample and feature collaborative clustering semi-supervised learning model is shown.The method can fully explore and utilize the bidirectional complex interaction mode of feature and sample and the label information of speech imagination in the research and system of speech imagination intention decoding based on electroencephalogram signal.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Semi-supervised collaborative clustering emotional electroencephalogram feature and sample coupling analysis method and device and storage medium of semi-supervised collaborative clustering emotional electroencephalogram feature and sample coupling analysis method

    CN118948297A