Papers, preprints, and technical reports.

EEG-Based Emotion and Depression Recognition Using a Hybrid GNN–Transformer Architecture

Rajeswari M, Santosh Srinivasaiah, Bhumika P Babu, S Geetha, Sanjitha Manya Raj, Likhitha J

A non-invasive BCI system that combines graph neural networks with transformer attention to recognize emotional states and assess depression levels directly from EEG. Tested on benchmark datasets, the hybrid architecture achieves competitive accuracy on both tasks — pointing toward affect-aware and mental-health-oriented applications.