Scientists Develop Child-Like Synthetic Voice for Children Who Can’t Speak
By Coxworth, Ben; Gizmag,Publication Date: February 21, 2012
Article reports on a new method of creating synthetic speech that sounds like that of a child. For its development, researchers in Norway created a master voice by combining recordings of multiple adult speakers reciting several thousand phrases, enough to create a workable library of words and sounds. Then a single child was recorded while reciting a smaller number of phrases selected to include the sounds that are most essential to the Norwegian language. A computer compared the master voice to the child’s voice, using the phrases as a point of reference, and was able to alter the master voice to sound like that of a child. The method is a result of a collaboration between scientists at software company Lingit and Media LT, a company that develops devices for assisted living. The childlike synthetic voice developed can be used in speech generating devices for children. Synthetic children’s speech has also been developed and used at the Norwegian University of Science and Technology to teach computer voice recognition systems to better understand children’s voices. Researchers at the university created a synthetic child’s voice by analyzing how children’s shorter vocal tracts affect the frequency distribution of their speech energy. They then altered the energy distribution of an adult speech program to achieve a childlike sound.
Published by: Gizmag Pty Ltd (Website:http://www.gizmag.com/)
Link to text: http://www.gizmag.com/childrens-synthetic-speech/21545/
