Per­haps the two most famous speech­less indi­vid­uals are physi­cist Stephen Hawking and the late Robert Ebert, a film critic. To com­mu­ni­cate with other indi­vid­uals, both men used com­puters that gen­er­ated syn­the­sized voices, which sounded tinny, stilted and unnatural.

Now new tech­nology could give speech­less people a nat­ural voice as unique to them as voices are to people who can speak.

Device Turns Thoughts Into Speech

Right now, people who need to use syn­the­sized voices to talk for them use a handful of generic voices, because cre­ating them is time-​​consuming and costly,” said Rupal Patel, Ph.D, asso­ciate pro­fessor at North­eastern Uni­ver­sity. “We feel strongly the voice from the device should reflect some­thing about that individual.”

With that phi­los­ophy, Patel, along with her research and devel­op­ment partner Tim Bun­nell, Ph.D., a pro­fessor at the Uni­ver­sity of Delaware, devel­oped VocaliD, a product that blends real human voices from healthy talkers with char­ac­ter­is­tics of the client’s unique speech pat­terns. The tech­nique, called voice mor­phing, enables Patel and Bun­nell to create a voice that is unique to the indi­vidual, and no longer has the now-​​traditional com­put­er­ized sound.

Read the article at Discovery News →