%0 Journal Article
%T 基于脑电信号和眼动信号的情感识别<br>Emotion Recognition Based on EEG Signals and Eye Movement Signals
%A 李澳
%A 张恒千
%A 尹钟
%J Modeling and Simulation
%P 1244-1252
%@ 2324-870X
%D 2025
%I Hans Publishing
%R 10.12677/mos.2025.141112
%X 本文探讨了基于脑电信号和眼动信号的多模态情感识别方法&#65292;使用了上海交通大学SEED-IV和SEED-V数据集。通过预处理脑电信号并提取功率谱密度和微分熵特征&#65292;结合眼动信号的时域和频域特征&#65292;提出了一种基于对比学习的数据融合方法&#65292;显著提高了情感识别的准确率。实验结果表明&#65292;多模态融合&#65292;尤其是脑电信号与眼动信号的结合&#65292;能显著提升识别性能&#65307;同时&#65292;微分熵特征优于功率谱密度特征&#65292;具有更好的分类准确性。<br>This paper explores a multimodal emotion recognition method based on EEG and eye movement signals, using the SEED-IV and SEED-V datasets from Shanghai Jiao Tong University. After preprocessing the EEG signals, power spectral density and differential entropy features were extracted, while time-domain and frequency-domain features were extracted from the eye movement signals. A contrastive learning-based data fusion method is proposed, which effectively bridges the distribution gap between different modalities and significantly improves emotion recognition accuracy. Experimental results show that multimodal fusion, particularly the combination of EEG and eye movement signals, significantly enhances recognition performance. Additionally, the differential entropy feature of EEG signals outperforms the power spectral density feature in classification accuracy, indicating better separability in the dataset.
%K 情感识别&#65292
%K 对比学习&#65292
%K 机器学习&#65292
%K 特征融合<br>Emotional Recognition
%K Comparative Learning
%K Machine Learning
%K Feature Fusion
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=106434