Emotion Recognition Using EEG and fNIRS Signals
Author Information
Author(s): Hou Mingxing, Zhang Xueying, Chen Guijun, Huang Lixia, Sun Ying
Primary Institution: Taiyuan University of Technology
Hypothesis
Can hybrid brain networks constructed from EEG and fNIRS signals improve emotion recognition performance?
Conclusion
The proposed method significantly enhances emotion recognition performance by integrating EEG and fNIRS signals.
Supporting Evidence
- The proposed method achieved recognition accuracies of 96.6% for SVM and 91.7% for KNN.
- Hybrid brain networks showed superior performance compared to unimodal approaches.
- Utilizing EEG source signals alleviated the volume conduction effect, enhancing recognition accuracy.
Takeaway
This study shows that combining two types of brain signals can help computers better understand how we feel.
Methodology
The study used EEG and fNIRS signals to construct hybrid brain networks for emotion recognition, employing Granger causality analysis and coupling strength calculations.
Limitations
The source localization process is time-consuming and only the neurovascular coupling relationship between EEG and fNIRS was investigated.
Participant Demographics
50 college students, 25 male and 25 female.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website