AINeutralarXiv – CS AI · Mar 44/104
🧠
MEBM-Phoneme: Multi-scale Enhanced BrainMagic for End-to-End MEG Phoneme Classification
Researchers developed MEBM-Phoneme, a neural decoder that uses magnetoencephalography (MEG) brain signals to classify phonemes with enhanced accuracy. The system integrates multi-scale convolutional modules and attention mechanisms to improve speech perception analysis from non-invasive brain recordings.