People with severe speech impairments have a serious communication problem, hearing and understand well, but not being able to respond coherently enough for others to comprehend what they’re saying. They’re not mute, but essentially pronounce words with a heavy unique accent. To help such folks keep up with conversations, researchers at University of Sheffield in the UK have developed a system that learns each disabled speaker’s unique voice characteristics and is able to perform voice recognition and transliterate what’s said into comprehensive synthesized speech.
Here’s a video report on the VIVOCA (Voice Input Voice Output Communication Aid) system:
VIVOCA project page: Voice Input Voice Output Communication Aid…
Press statement from the University of Sheffield: Yorkshire technology bestows the power of speech…