EEG-based Emotion Recognition Using Spectral Graph Convolutional Neural Network based on Partial Directed Coherence
Oral presentation
Research approach
Constructed EEG channel relationships with Partial Directed Coherence and classified emotions with a spectral GCN.
Results
Subject-dependent SEED evaluation: SGCN-PDC accuracy 90.93 ± 4.86%, compared with 90.40 ± 8.49% for DGCNN.
All publications