American poet Henry Wadsworth Longfellow once said, “The human voice is the organ of the soul.” It’s also been said the human voice is not just a sound but also a reflection of the person. Everyone has a unique voiceprint, yet the millions of individuals who rely on communication devices have been given uniform, robotic voices that don’t fit their personalities or bodies. VocaliD is hoping to bid adieu to these synthetic voices by using cutting-edge technology to create personalized voices for these individuals.
The creator behind VocaliD, Rupal Patel, stated, “We wouldn’t dream of fitting a little girl with the prosthetic limb of a grown man, so, why then the same prosthetic voice?” So how exactly is the VocaliD voice achieved? It is obtained through a two-part process which involves the recording and blending of voices. According to its Indiegogo website, here is how a VocaliD voice is obtained:
Through years of research we observed that even those who have severe speech impairment produce distinctive sounds when they laugh, cry, and vocalize emotions and intentions. Their vocalizations contain the voice’s source—its pitch, tone, rhythm, and loudness. What these individuals lack is the ability to manipulate their tongue, lips, and the shape of their mouth—their vocal filter. The filter is essential to producing clearly articulated speech. So we developed technology to harness the source characteristics of recipients and supplement them with the vocal filters available to healthy talkers. The result is a blended voice that is as understandable as the donor while conveying the vocal identity of the recipient.
VocaliD has a three-part mission:
- Banking Speech: building the Human Voicebank, the world’s largest, most diverse collection of speech.
- Blending Voices: leveraging cutting-edge technology to craft voices from donor and recipient speech sounds.
- Empowering Lives: helping people engage in the world, speak as themselves and have their own voices heard.