Neuroscientists at Columbia University have created a system that translates thoughts into speech. By monitoring brain activity, technology can reconstruct words that a person hears, with unprecedented clarity. This may lead to new ways for computers to communicate directly with the brain.
It can also help people who can’t speak, such as those with amyotrophic lateral sclerosis or those recovering from a stroke, Tech Xplore writes.
The discovery was described in detail in the new issue of Scientific Reports published on January 29, 2019.
“Our voices help us to establish relationships with friends and family, which is why the loss of speaking ability is so devastating,” said Nima Mesgarani, lead author of the study. “With the help of this study, we have the opportunity to restore this capacity. I showed that with the right technology, the thoughts of these people can be decoded and understood by any listener,” added the researcher.
This was difficult to achieve. The first efforts to decode brain signals by Mesgarani and others focused on simple computer models that analyzed spectrograms – visual representations of sound frequencies. But because this approach failed to produce intelligible speech, the team used a computer algorithm that synthesizes speech after being trained with speech recordings.
“It’s the same technology used by Amazon’s Echo and Apple’s Siri that provide verbal answers to our questions,” Mesgarani said.
Although the project is in an incipient phase, with only simple words and expressions being rendered, the team plans to test the accuracy of more complicated sentences and words. Ultimately, the researchers hope their system will be part of an implant that translates the bearer’s thoughts directly into words.