OpenAI Whisper - CPU Application of PyTorch Dynamic Quantization for CPU deployment to OpenAI Whisper model with benchmarks
DeepSpeech2 LJSpeech and sEMG Silent Speech classification
EEG/sEMG Speech
sEMG Silent Speech - Speech Recognition State-of-the-art (SOTA) approach for sEMG silent speech ASR which builds on the SOTA transduction model for silent speech
kara-one-transduction Experiments demonstrating
feasibility of synthesizing speech from EEG signals during stimuli, vocal and imagined conditions. Applying SOTA
sEMG Silent Speech synthesis methods to EEG signals is viable.