Progressive graph convolution network for EEG emotion recognition

• Published in: Neurocomputing (Amsterdam) Vol. 544; p. 126262
• Main Authors: Zhou, Yijin, Li, Fu, Li, Yang, Ji, Youshuo, Shi, Guangming, Zheng, Wenming, Zhang, Lijian, Chen, Yuanfang, Cheng, Rui
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2023
• Subjects: Brain region; EEG emotion recognition; Emotional hierarchical categories; Progressive graph convolution network (PGCN); EEG emotion recognition; Brain region; Emotional hierarchical categories; Progressive graph convolution network (PGCN)
• ISSN: 0925-2312; 1872-8286
• DOI: 10.1016/j.neucom.2023.126262

Studies in the area of neuroscience have revealed the relationship between emotional patterns and brain functional regions, demonstrating that the dynamic relationship between different brain regions is an essential factor affecting emotion recognition determined through electroencephalography (EEG). Moreover, in EEG emotion recognition, we can observe that clearer boundaries exist between coarse-grained emotions than those between fine-grained emotions, based on the same EEG data; this indicates the concurrence of large coarse- and small fine-grained emotion variations. The progressive classification process from coarse- to fine-grained categories may be helpful for EEG emotion recognition. Consequently, in this study, we proposed a progressive graph convolution network (PGCN) for capturing this inherent characteristic in EEG emotional signals and progressively learning the discriminative EEG features. To fit different EEG patterns, we constructed a dual-graph module to characterize the intrinsic relationship between different EEG channels, containing the dynamic functional connections and static spatial proximity information of brain regions from neuroscience research. Moreover, motivated by the observation of the relationship between coarse- and fine-grained emotions, we adopted a dual-head module that enabled the PGCN to progressively learn more discriminative EEG features, from coarse-grained (easy) to fine-grained categories (difficult), referring to the hierarchical characteristics of emotion. To verify the performance of our model, extensive experiments are conducted on three public datasets: SEED-IV, SEED-V, and MPED. The experiment results show that the PGCN achieves a state-of-the-art performance. Furthermore, we explored the effect of different frequency bands based on our model and visualized the activated brain regions. The experiment results reveal the relationship between human emotion and high-frequency EEG signals, as well as the importance of the frontal and temporal lobes for emotion expression.