Daily US Times: Scientists have taken a step forward in their ability to understand what a person is saying just looking at their brainwaves. They trained algorithms which can transfer brain patterns injto sentences with word error rates as low as 3%.
A new study published in the journal Nature Neuroscience revealed the success.
Scientists’ had limited success in decoding neural activity previously.
The earlier efforts in this area were only able to understand small amount of spoken words or a small percentage of the words contained in particular phrases.
Dr Joseph Makin, who is a Machine learning specialist from the University of California, San Francisco (UCSF), and his colleagues tried to improve the accuracy.
Four volunteers read sentences aloud while electrodes recorded their brainwaves. Then, the brain activity was fed into a computing system, which created a representation of regularly occurring features in that data.
These patterns are likely to be related to repeated features of speech such as commands, vowels or consonants to parts of the mouth.
Another part of the system decoded this representation word-by-word to form sentences.
The authors freely admit the study’s caveats though, such as the speech to be decoded was limited to 30-50 sentences.
The researchers wrote in their Nature Neuroscience paper: “Although we should like the decoder to learn and exploit the regularities of the language, it remains to show how much data would be required to expand from our tiny languages to a more general form of English.”
But they add that the performance was improved by adding sentences to the training set that were not used in the tests themselves.
You may read: Mysterious radio signal from space repeats every 16 days