基于多域解耦的脑电情绪识别方法、电子设备及介质
By employing a multi-domain decoupled EEG emotion recognition method, which utilizes variational autoencoders and emotion classifiers to decouple EEG signals, cross-individual and cross-group emotion recognition can be achieved without target domain data. This solves the problems of individual difference interference and poor cross-group adaptability in existing technologies, and improves the stability and accuracy of recognition.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHUNTONG INFORMATION TECH (DALIAN) CO LTD
- Filing Date
- 2026-05-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing EEG signal emotion recognition methods suffer from significant individual differences, poor cross-individual generalization ability, deep coupling between individual identity information and emotional information with incomplete decoupling, poor cross-group adaptability, and insufficient model feature discriminativeness and robustness, making it impossible to achieve zero-sample cross-individual and cross-group emotion recognition.
A multi-domain decoupled EEG emotion recognition method is adopted. By constructing a variational autoencoder, a domain discriminator, and an emotion classifier, the EEG signal is decoupled into individual identity, emotional state, and noise features. KL divergence, mutual information minimization, and orthogonal constraints are used to force feature independence. Combined with adversarial training and group adaptive regularization, cross-individual and cross-group emotion recognition can be achieved without target domain data.
It effectively removes individual redundancy and noise interference in emotional features, improves the stability and reliability of emotion recognition, lowers the application threshold, achieves cross-group adaptation, improves recognition accuracy and cross-scenario generalization performance, and adapts to emotion assessment and monitoring scenarios.
Smart Images
Figure CN122182044B_ABST