Focus on Technology
2:00 pm
Wed April 16, 2014

A voice bank for the speechless

An effort is underway to give individuality to people who have to use a computerized voice box to speak.

Just in the United States alone there are several million people who have lost their voice because of cerebral palsy, stroke, brain injury and more. If they are lucky enough to have a computerized voice chances are their voice sounds very similar to somebody else who also uses a computer.

Former Jazz host and WVXU contributor David Delegator saw this TED speech by speech scientist Dr. Rupal Patel of Northeastern University and decided to donate his voice.

Delegator has signed up to become part of the Human Voice Bank Initiative VocaliD.

He says he wants to help the speechless gain an identity. "You may have heard somebody like Stephen Hawking, the physicist and thinker. He's incapable of speech but he has this... awful machine voice and people psychologically don't identify with machine voices."

How it works

  1. Donors sign up here.
  2. Recipients sign up here.
  3. Recipients sometimes have some speech left. (vowels and consonants) This is recorded.
  4. Donors will record about 3,000 sentences in 2-3 hours.  This is not done in one sitting. It will be a "game" in the form of an iPhone or web app.
  5. Dr. Patel blends the two voices together.

Listen to how the voices sound blended.

In the next few months Dr. Patel will be ready for the donors to record their voices. Right now she's in the process of developing new technology to collect them.