Domain adaptation spatial feature perception neural network for cross-subject EEG emotion recognition

• Published in: Frontiers in human neuroscience Vol. 18; p. 1471634
• Main Authors: Lu, Wei, Zhang, Xiaobo, Xia, Lingnan, Ma, Hua, Tan, Tien-Ping
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Media S.A 17.12.2024
• Subjects: affective computing; convolutional neural network; domain adaptation; electroencephalography; emotion recognition; graph attention network; Human Neuroscience; graph attention network; electroencephalography; domain adaptation; affective computing; emotion recognition; convolutional neural network
• ISSN: 1662-5161; 1662-5161
• DOI: 10.3389/fnhum.2024.1471634

Emotion recognition is a critical research topic within affective computing, with potential applications across various domains. Currently, EEG-based emotion recognition, utilizing deep learning frameworks, has been effectively applied and achieved commendable performance. However, existing deep learning-based models face challenges in capturing both the spatial activity features and spatial topology features of EEG signals simultaneously. To address this challenge, a d omain-adaptation s patial-feature p erception-network has been proposed for cross-subject EEG emotion recognition tasks, named DSP-EmotionNet. Firstly, a s patial a ctivity t opological f eature e xtractor m odule has been designed to capture spatial activity features and spatial topology features of EEG signals, named SATFEM. Then, using SATFEM as the feature extractor, DSP-EmotionNet has been designed, significantly improving the accuracy of the model in cross-subject EEG emotion recognition tasks. The proposed model surpasses state-of-the-art methods in cross-subject EEG emotion recognition tasks, achieving an average recognition accuracy of 82.5% on the SEED dataset and 65.9% on the SEED-IV dataset.