Title: Research of EEG-based emotion recognition for the deaf with feature fusion

Authors: Zemin Mao; Xuewen Zhao; Yu Song

Addresses: Technical College for the Deaf, Tianjin University of Technology, Tianjin 300384, China ' School of Electrical Engineering and Automation, Tianjin University of Technology, Tianjin 300384, China ' School of Electrical Engineering and Automation, Tianjin University of Technology, Tianjin 300384, China

Abstract: Electroencephalogram (EEG) is better at reflecting emotional changes. The paper aims to explore the deaf brain activity mechanisms of emotional processing with EEG signals. Fifteen deaf subjects were recruited to participate in the emotional induction experiment, and EEG signals were collected when they watched three kinds of emotional movie clips. The frequency domain and brain network features were extracted and fused to capture correlation among EEG channels. The results show that fused features outperform single differential entropy (DE) features in classification indicators (accuracy: 96. 73%, recall: 96.58%, precision: 97.36%, F1 score: 96.42%). In addition, a stacking ensemble learning framework was proposed to classify the fused features, achieving higher classification accuracy than SVM by 2.72%. Investigation into brain activities reveals that deaf brain activity changes mainly in the beta and gamma bands, and the brain regions that are affected by emotions are distributed primarily in the frontal and occipital lobes.

Keywords: emotion recognition; deaf subject; electroencephalogram; EEG; brain network; differential entropy; DE.

DOI: 10.1504/IJBET.2024.138973

International Journal of Biomedical Engineering and Technology, 2024 Vol.45 No.3, pp.216 - 236

Received: 07 Jun 2023
Accepted: 22 Sep 2023
Published online: 05 Jun 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article