Thanks to its My Own Voice application, the Toulouse company Acapela gave back his voice to a patient who was unable to speak following removal of the vocal cords and tongue.
Giving back their voice to those who have lost it following accidents or health problems. Such is the feat that the company Acapela, which specializes in digital voices, has managed to accomplish. Thanks to the artificial intelligence of its My Own Voice appthe Toulouse company has just given back his voice to Maurice Foissac, a resident of Occitania unable to speak following the removal of the vocal cords and tongue.
Using an algorithm, Acapela reproduces the patient’s voice print by making him record 350 specific phrases designed by linguists. “A few days before his operation, Maurice recorded his voice one last time”, explains France 3 in a report on the subject. This process took an hour. Two days later, his voice, his intonations, his Southern accent, had been translated into synthetic voice.
Today, the vocal exchange is done thanks to a tablet where the patient writes then everything is transcribed with his own voice. “The added value of this solution is above all to be able to speak when there are several of us discussing. On the other hand, the others must be silent when I speak”, explains Maurice to France 3.
A solution little known in France
“We have work in progress on artificial intelligence which makes it possible to create new synthetic voices more quickly and easily. This work allows us at the same time to improve the quality and to imitate the more and more faithful human speech”, explains the company.
“The voice is important for a person and their loved ones. This poses the problem of voice identity in a world where people communicate more and more with connected objects. Our solution is little known in France because the health system is centralized towards the institutions, unlike the United States where we are solicited by the associations”, specifies Rémy Cadic, president of Acapela, quoted by 20 minutes.
“We also work with children, in particular autistic people, to allow them to express themselves. For these voices, we favor children’s words so that they keep their child identity to talk with their friends”, adds Cédric Cadic .
A service available in 13 different languages
If voice synthesizers have been used for years, the particularity of My Own Voice “is to reconstruct the person’s voice. It’s very original and interesting”, explains Maurice Foissac’s doctor. “The service does not require any special skills or equipment, except for a computer, a headset and an Internet connection”, argues the company on its website.
The service, available in 13 different languages, is free from creation to testing, which can last up to three months, it is specified. Once the customer is satisfied with the result, he will have to pay 999 euros to use it unlimited.
In addition to its work in accessibility, Acapela has been providing the voices that assist our daily journeys (bus, metro, Toulouse tram, GPS), voice servers or assistant robots for thirty years. These digital voices are translated into 30 languages around the world.
.
