Neuro-engineers at Columbia University in New York state that they have created a system that can translate human thought into a recognizable speech that would revolutionize not only medicine but also communication.
By observing the brain activity of the subjects, researchers at the Columbia Brain Brain Institute, Mortimer B. Zukerman, managed to teach artificial intelligence to translate thought patterns into understandable sentences, says a report published on Tuesday in the journal Scientific Reports, . The authors see patients with speech-disturbed disease or trauma as the first adopters of the emerging technology.
"We have shown that with the right technology the thoughts of these people can be decoded and understood by every listener" According to Dr. Nima Mesgarani, the senior author of the newspaper, after early researchers' attempts to translate brain activity in a recognizable speech, they turned to a computer algorithm that could generate a speech called a vocoder. The algorithm improves what is "trained" from records of human speech.
Researchers translate directly into the brain signal speech: //t.co/qEStGOoPOW
This breakthrough, which uses the power of speech synthesizers and artificial intelligence, can lead to new ways of direct communication between computers and the brain. #BCI #speech #science
̵1; Neuroscience News (@NeuroscienceNew) January 29, 2019
"This is the same technology used from Amazon Echo and Apple Siri to give oral answers to our questions, said Dr. Messgani, who is also a professor at the Columbia School of Engineering and Applied Sciences, Fu Food
The vocoder was trained to interpret the brain with Dr. Ashes Dinesh Mehta, a neurosurgeon at the Northwell Health Neuroscience Institute on Long Island, and the co-author of the paper. "Working with Dr. Mehta, we asked patients with epilepsy, who have already been subjected to brain surgery, to listen to sentences pronounced by various said Dr. Messgani. "These neural models train the vocoder."
Once this training has been conducted, the next phase begins. Patients listened to a person reading numbers 0 through 9 until the algorithm scanned the brain activity and tried to make it sound. The result is a robotic voice for reading numbers that human listeners can understand and repeat with about 75 percent accuracy.
This may seem rather modest, but Dr. Messgani said that such a result is above and beyond all previous Researchers are planning to improve the system further so that they can take as input brain patterns who thinks about speech, but does not actually listen to it. This will give anyone who has lost his ability to speak, either through injury or illness, a renewed chance to connect with the world around them " Dr. Messgani said.
Technology will also have to work with more complex words and sentences to make it more practical. The ultimate goal of the team is to create an implant to synthesize the speech directly from the thought.
Do you think your friends will be interested? Share this story!