Cross-Subject and Cross-Montage EEG Transfer Learning via Individual Tangent Space Alignment and Spatial-Riemannian Feature Fusion
The study introduces Individual Tangent Space Alignment (ITSA), a pre-alignment strategy combining subject-specific recentering, distribution matching, and supervised rotational alignment to improve cross-subject generalization in EEG transfer learning. This method fuses Regularised Common Spatial Patterns with Riemannian geometry to enhance feature extraction for gait rehabilitation BCIs.