Giving CANDY to Children: User-Tailored Gesture Input Driving An Articulator Based Speech Synthesizer

Authors:Pausch, Randy, Department of Computer ScienceUniversity of Virginia Williams, R, Department of Computer ScienceUniversity of Virginia

The CANDY Project (Communication Assistance to Negate Disabilities in Youth) seeks to provide a realtime speech synthesizer for disabled individuals, particularly non-vocal children with cerebral palsy. Existing speech synthesizers convert user input into discrete linguistic or phonetic symbols which are converted into sound. Complicated sentences must be created by concatenating lower level symbols, precluding real-time conversational speech. We have developed an articulator-based speech synthesizer which simulates the motion of the human tongue and produces the corresponding speech sounds in real time. The synthesizer is driven by two continuous input signals and non-disabled users can produce realtime speech with a joystick. Disabled users will drive the synthesizer via passive tracking of their body movements. Magnetic trackers attached to the user report their location and tailoring software allows each user to move the tracker in an optimal orientation and range. The user motion is then converted into the two continuous signals that drive the speech synthesizer. In this way, we hope to allow each child to compensate for their inoperative vocal tract by using their "best" set of muscles to operate a simulated vocal tract. The motion mapping software may also have future potential as a physical therapy aid. to appear in: Communications of the ACM
Note: Abstract extracted from PDF text

All rights reserved (no additional license for public reuse)
Source Citation:

Pausch, Randy, and R Williams. "Giving CANDY to Children: User-Tailored Gesture Input Driving An Articulator Based Speech Synthesizer." University of Virginia Dept. of Computer Science Tech Report (1991).

University of Virginia, Department of Computer Science
Published Date: