Using new techniques, a computer was able to recreate sound a person heard just by scanning their brain. See here a video report from Newsy:
Embedded Video Source by Newsy.com
Transcript by Newsy
BY STEVEN SPARKMAN
ANCHOR LAUREN ZIMA
Researchers are one step closer to hacking into the human brain. New techniques from UC Berkeley let a computer listen to a conversation — by scanning the brain. KPIX explains.
“They’re decoding the electrical activity that happens when a patient hears a certain word. … Researchers then attempt to reverse the process: take the brain activity and then have a computer model say the word.”
The researchers read patients lists of words while measuring how the brain lights up. After the program learned its way around a particular brain, it could reproduce the sounds patients were hearing, even if the program hadn’t learned that word before.
(Computer sounds) (Video source: UC Berkeley)
The sound was muddled, but analysts say it’s a proof of concept, and shows that scientists are getting ever closer to decoding the brain.
Last year, researchers did something similar with vision. Based on the brain’s signals, a computer model was able to reconstruct rough images of what a person was seeing. (Video credit: Gallant Lab)
So where is all this going? Scientists think the same thing happens in the brain when a person hears a word as when a person thinks of a word. A writer for Popular Science explains, if that idea holds up, these techniques could give a new voice to those who have lost theirs.
“[T]here are some practical, medically motivated reasons to do these things, like communicating with locked-in patients, or those who have lost the ability to speak because of a stroke or a degenerative muscle disease.”
The researchers caution it’ll be decades before devices that can translate thought to speech exist. And what about people tapping into your brain without an invitation? The researchers say that’s even more remote. An analyst on BBC says the program needs your help to work at all.
“There’s no need to go out and buy a tinfoil hat just yet. Again, this is very invasive, this required a willing participant to sort of teach the computer model how their particular brain gives up the signals and so on.”
Transcript by Newsy.
(Image source: UC Berkeley)