GL-LFGNN:A Global-Local Dual-branch Causal Graph Neural Network Based on Liang-Kleeman Information Flow for EEG Emotion Recognition
Researchers have developed a novel Graph Neural Network (GNN) called GL-LFGNN for recognizing emotions from EEG data. This model utilizes the Liang-Kleeman information flow theory to capture causal influences in neural activity, moving beyond traditional statistical associations. The dual-branch architecture integrates whole-brain connectivity with region-specific processing, achieving high accuracy on the MEEG dataset with significantly fewer parameters than existing methods. AI
IMPACT Introduces a new causal modeling approach for EEG analysis, potentially improving accuracy and efficiency in emotion recognition.