VocaliD offers customized synthetic voices for individuals unable to speak
Individuals who are unable to communicate using their own voices may soon have the option to speak through customized synthetic voices instead of using the robotic, GPS-like electronic voices that are used today.
Researchers at VocaliD – a technology company based in Boston, Massachusetts in the US – are currently developing a first-of-its-kind speech technology that captures the spirit of an individual’s personality. Their aim is to address the frustration felt by millions with severe speech disorders who must communicate through computerized devices.
“When you lose your voice and you forever have to communicate through, let’s say, a GPS-sounding voice, you’re losing a piece of your identity,” said speech scientist and VocaliD founder Rupal Patel, a speech technology professor on leave from Northeastern University.
“Our technology is basically swapping out the generic voice for a voice that is made from a person’s own little bit of sound,” Patel added.
People who are speechless still make sounds. VocaliD’s speech technology harnesses those sound patterns and mixes them with a real human voice. The donor voice comes from someone who is similar in age, gender and accent. The goal is to give a voice to a person who can’t speak — no two voices are alike.
Some of the donor voices come from middle-school students in California. Instead of a traditional fundraiser or a bake sale, the teenagers decided as a class project to donate their time and, most importantly, their voices to the Human Voice Bank, which is part of VocaliD.
The volunteers spend a couple of afternoon and weekend sessions recording their voices, which then end up in a collection of almost 19,000 voices from around the world.
Subscribe
If you enjoyed this article, subscribe to receive more just like it.
