Group Resonance Network Enhances EEG Emotion Recognition

The Group Resonance Network (GRN) introduces a novel approach to improving emotion recognition from electroencephalography (EEG) signals by addressing inter-subject variability. This method utilizes individual EEG dynamics alongside group resonance modeling to enhance classification accuracy. The GRN architecture comprises key components including an individual encoder, learnable group prototypes, and a specialized multi-subject resonance branch, which collectively facilitate more robust emotion recognition across different subjects.