NEW TECHNOLOGY MADE SPEACHLESS EMPOWERED
A new technology that could help speechless people who use devices to talk, speak in real human voices instead of a computerised sound, has been developed by researchers, including one of Indian-origin. One of the most famous speechless individuals is British physicist Stephen Hawking. To communicate with people, he uses computers that generate synthesised voices, but they sound stilted and unnatural. Scientists have now developed VocaliD, a product that blends real human voices from healthy talkers with characteristics of the client's unique speech patterns. "Right now, people who need to use synthesised voices to talk for them use a handful of generic voices, because creating them is time-consuming and costly," said Rupal Patel, associate professor at Northeastern University. "We feel strongly the voice from the device should reflect something about that individual," Patel said. Patel, along with Tim Bunnell, a professor at the University of Delaware developed VocaliD, which uses a technique called voice morphing, 'Discovery News' reported. The technology has positive implications for individuals who are autistic, or who suffer from disorders such as Amyotrophic lateral sclerosis (ALS) or stroke, Patel said. "There's a relatively broad market of two-and-a-half to three million people who use devices to talk for them," said Patel. "We'd love to see this technology being available to that population. We feel that if we can personalise the device, they will be more likely to use it and more socially acceptable," she said. The technology is based on the fact that even speechless individuals can still make sounds, Bunnell said. "We can always grab characteristics of their voice and reapply the process as they go along," he said. "With individuals who have neuro-degenerative diseases such as ALS, we capture speech from them right after they're diagnosed, while they're still speaking fluently and create their new voice from that. It captures elements of their voice that default devices don't typically use," he added. VocaliD was developed as an extension of another product called Model Talker, a text-to-speech synthesis system developed by Bunnell. VocaliD aims to personalise the speech so that all vowels and consonants are clearly understandable.