Silent Speech

Communication technologies that generate intelligible speech from non-acoustic signals produced during the intent or attempt to speak, without requiring audible voicing. Input modalities include surface electromyography of articulatory muscles, ultrasound tongue imaging, lip-reading from cameras, real-time MRI, non-audible murmurs captured by stethoscopes or in-ear sensors, and bone conduction. Silent speech interfaces are motivated by accessibility (users who cannot produce voiced speech, such as those with laryngectomy or severe dysarthria) and by privacy or quiet-environment use. Challenges include hardware portability, restricted vocabularies, and robust open-vocabulary decoding.