Frequency-Decorrelated Temporal Ensembles for EEG--fNIRS Imagined-Handwriting Decoding
FRED is proposed, a task-adapted system that models imagined handwriting as a multi-second motor sequence and trains a compact multi-scale temporal network on three complementary EEG frequency views and identifies frequency-diverse temporal EEG modeling and protocol-matched structured inference as the principal sources of performance in this sparse-montage EEG--fNIRS setting.