[INTERSPEECH 2025]Official code for "MiSTR: Multi-Modal iEEG-to-Speech Synthesis with Transformer-Based Prosody Prediction and Neural Phase Reconstruction"
speech-synthesis transformer brain-computer-interface multimodal-learning mel-spectrogram ieeg neural-vocoder phase-reconstruction neural-decoding bio-signal-processing brain-to-speech prosody-modeling
-
Updated
May 31, 2025