Teams of scientists from the National Center for Scientific Research (CNRS) and the BrainTech Laboratory of the National Institute for Health and Medical Research (Inserm) have designed a speech synthesizer producing intelligible speech.
Device translates silent articulation into speech
His function ? Recreate the voice of a person who is not able to speak (dumb, paralyzed and aphasic). In practice, it suffices to move the tongue, the lips and jaw.
A learning algorithm, developed beforehand, makes it possible to decode these silent articulatory movements thanks to sensors placed on the tongue, lips and jaw. Finally, the system verbalizes them in real time in synthetic speech.
A future interface to restore speech from the brain
Scientists explain in the journal PLOS Computational Biology that this device can be used by any person, after a short period of calibration. There would be no vocabulary restriction.
Next step: develop a brain-computer interface allowing the speech restoration from the sole activity of the brain.
Read also :
Video: a glove that makes deaf-mutes talk