Human Voice Bank: New Technology Gives Chance To Those Who Can’t Speak
The human voice bank has attracted numerous volunteers to donate their voices to people who rely on technology to communicate. Currently, there are computerized devices that sound robotic and do not match the person's identity.
VocaliD, a company based in Boston, aims to address the need of those with severe speech disorders. Rupal Patel, a speech scientist and VocaliD founder, believes that voice is an integral part of a person's identity. He added that the human voice bank is swapping the generic voice for a sound from the person who has a speech problem.
The new human voice bank harnesses sound patterns from people who are speechless, and mixes them with a real human voice. The donor voice is matched with the recipient according to accent, age and gender.
The human voice bank is becoming popular attracting 17,000 volunteers from 110 countries. In America alone, there are two million who suffer from severe speech disorders. Students from different universities who chose to take part in this advocacy have begun donating their time and voice. Instead of the usual fundraising activity, they spend an afternoon or the weekend recording their voices.
Another company, Speech Research Laboratory at the Alfred I. Dupont Hospital for Children in Wilmington, Delaware, headed by Tim Bunnell, has also engineered more than 1,000 voices for people with amyotrophic lateral sclerosis. Unlike the VocaliD human voice bank, they have recorded voices for patients who were able to record their own before they lost their speech ability.
A hundred words takes about 20 minutes to record, and volunteers are asked to do 3,000 sentences, details a 7th grader donor in California. VocaliD will also record the sounds of the users and blend them with the donor voice for about six to ten hours of recording. The human voice bank is reported to have collected 19,000 voices around the world from donors aging six to 91 years old.