Scientists have created an implant to turn thoughts into voice

Neurologists from the University of California at San Francisco have created an implant that can decode brain signals and convert them into synthesized speech. On the basis of the development can be built neural speech prostheses. In the future, it will help to communicate normally to people suffering from stroke, head injuries and various neurodegenerative diseases such as Parkinson's disease, multiple sclerosis and amyotrophic lateral sclerosis.

“For the first time in history, we can generate whole oral sentences based on human brain activity. This is proof that, with the technology already available, we can create a device that is clinically viable in patients with speech loss, ”said Edward Chang, a professor of neurological surgery and an employee of the Weil Institute of Neurobiology.

The designed implant is placed on the patient's brain and scans the signals. Then he decodes the received information and transmits it to another device that generates synthetic speech.

As Edward Chang said, accurate comparison of sound with anatomy allows scientists to create a realistic virtual voice for each patient. To do this, they used two machine learning algorithms. The first is used for a decoder that converts brain activity signals generated during speech into movements of the virtual vocal tract. The second algorithm is used by a synthesizer that converts the received movements of the vocal tract into a synthetic voice of the patient.

4pda.ru