According to foreign media New Atlas on the 14th, brain-computer interface technology has made a major breakthrough: a new implant system developed by the University of California, Davis, allows a paralyzed patient to not only speak smoothly, but also "hum" the melody with the help of a computer, with almost no delay.

The research team tested this technology on an ALS patient. By implanting four sets of microelectrode arrays in the area of ​​the brain that controls pronunciation, the system can capture the neural signals generated when speaking in real time, and then combine low-latency processing and AI decoding models to synchronously convert these signals into speech and play them in the speaker.

It should be emphasized that this technology is not "reading the mind", but identifying the motor commands issued by the brain when the patient tries to speak. The sound synthesized by the system is very similar to the patient himself - this is thanks to the sound cloning algorithm trained with the voice samples recorded before the onset of the disease.

From signal acquisition to speech generation, the whole process only takes 10 milliseconds, which can almost achieve real-time conversation.

The system can also recognize the patient's attempt to sing, determine the pitch he wants to express, and synthesize the melody accordingly. In the video released by the research team, this function is still rudimentary, but it is undoubtedly a hope for regaining voice for those who are gradually losing the ability to express themselves.

Sergey Stavisky, senior author of the research paper, said that this progress has brought a turning point for brain-computer speech synthesis. He said: "With instant voice output, neural prosthetic users can truly participate in conversations - they can interrupt others and are less likely to be interrupted by others by mistake."

The system can reproduce interjections such as "ah", "oh" and "um", determine whether the sentence is a question or a statement, and even identify the position of emphasized words. The research team also pointed out that the system can accurately synthesize fictional words that have never been seen before. This makes voice expression more natural and expressive.

According to the report, the researchers also pointed out that the current trial is only conducted on one patient, and the next step will be to expand to other people who have lost their language ability for different reasons to verify the universality of the technology.