PGCN: Pyramidal Graph Convolutional Network for EEG Emotion Recognition

TL;DR

Emotion recognition from EEG signals benefits from modeling inter-electrode relationships across multiple scales. The authors propose PGCN, a pyramidal graph convolutional network that integrates local (distance-based), mesoscopic (region-based with virtual centers), and global (attention-based) connections to capture short-, mid-, and long-range dependencies while mitigating over-smoothing. Key contributions include a learnable sparse local adjacency with inverse-distance weighting, two mesoscopic partition schemes with virtual centers, and a global attention-guided graph using position-augmented nodes, all fused for final classification; experiments on SEED, SEED-IV, and SEED-V show state-of-the-art performance in both subject-dependent and subject-independent settings. The study demonstrates improved representation richness and robustness to cross-subject variability, with visualization analyses illustrating region-specific activation patterns and connections tied to emotion processing.

Abstract

Emotion recognition is essential in the diagnosis and rehabilitation of various mental diseases. In the last decade, electroencephalogram (EEG)-based emotion recognition has been intensively investigated due to its prominative accuracy and reliability, and graph convolutional network (GCN) has become a mainstream model to decode emotions from EEG signals. However, the electrode relationship, especially long-range electrode dependencies across the scalp, may be underutilized by GCNs, although such relationships have been proven to be important in emotion recognition. The small receptive field makes shallow GCNs only aggregate local nodes. On the other hand, stacking too many layers leads to over-smoothing. To solve these problems, we propose the pyramidal graph convolutional network (PGCN), which aggregates features at three levels: local, mesoscopic, and global. First, we construct a vanilla GCN based on the 3D topological relationships of electrodes, which is used to integrate two-order local features; Second, we construct several mesoscopic brain regions based on priori knowledge and employ mesoscopic attention to sequentially calculate the virtual mesoscopic centers to focus on the functional connections of mesoscopic brain regions; Finally, we fuse the node features and their 3D positions to construct a numerical relationship adjacency matrix to integrate structural and functional connections from the global perspective. Experimental results on three public datasets indicate that PGCN enhances the relationship modelling across the scalp and achieves state-of-the-art performance in both subject-dependent and subject-independent scenarios. Meanwhile, PGCN makes an effective trade-off between enhancing network depth and receptive fields while suppressing the ensuing over-smoothing. Our codes are publicly accessible at https://github.com/Jinminbox/PGCN.

Figures (8)

• Figure 1: The EEG emotion recognition paradigm and the conceptual design of PGCN. (a) The basic flowchart of EEG emotion recognition, which consists of visual-audio stimuli presentation (to evoke corresponding emotions), EEG signal acquisition, pre-processing, feature extraction, emotion classification and (sometimes) feedback to the subject. (b) There are three main components for feature aggregation in the PGCN. The vanilla GCN extracts local bias information from neighboring nodes, the virtual mesoscopic center aggregates information within brain regions constructed with priori knowledge guidance, and the attentional GCN further fuses structural and functional connectivity at a global level.
• Figure 2: The flowchart of the proposed PGCN. To excavate the electrode relationships, PGCN aggregates multiscale information, i.e., local, mesoscopic and global features to conduct the emotion classification task. GCNs are used to model the relationships.
• Figure 3: Brain region division based on priori knowledge.
• Figure 4: Subject-dependent emotion recognition accuracy on the SEED-IV dataset.
• Figure 5: The curve of node smoothness with the increasing number of network layers. Node smoothness calculates the mean of the cosine similarity between the nodes of the output features of each layer.