Scientists have developed an artificial intelligence tool that can help people communicate without using their voicebox.

The technology analyses how the mouth is moving and then turns this into a robot voice, making it perfect for people who may have had throat surgery or other conditions that means their voicebox doesn’t work.

Developed by scientists at the Centre National de la Recherche Scientifique in Grenoble created the interface to help those with vocal cord paralysis.

‘This synthesiser converts movements of the main speech articulators (tongue, jaw, velum, and lips) into intelligible speech,’ the researchers wrote in the study.

‘Then computational models based on DNNs were trained on these data to transform articulatory signals into acoustic speech signals.”

Because the technology behind the robot voicebox is so sophisticated, using machine learning to work out what the person is saying, it can recognise very complex patterns, with little room for error at all.

‘We further assessed the real-time control of the synthesiser by both the reference speaker and new speakers,’ the authors said, ‘Speech Brain-Computer-Interfaces may restore communication for people with severe paralysis by decoding cortical speech activity to control in real time a speech synthesiser,’ the authors said.

The next step in the computer’s development is to completely bypass the mouth signals and translate the brain signals into speech, effectively predicting what the person will say. This could help those who have complete body paralysis speak again – groundbreaking tech that has the potential to help millions of people.
Although this is the first demonstration of using AI to work out what someone is saying, it’s not the first time scientists have used AI with speech. Google has already launched its algorithm that can translate seamlessly between languages using machine learning to understand the context of speech.