Auditory-Visual Speech Processing
We describe recent progress we have made in facial animation and visible speech synthesis. We have added new internal structures (three-dimensional teeth, hard palate, and higher resolution tongue with a larger number of controls). These structures and the outside of the face are being trained on measurements of real faces, using optotrak, electropalatography, and ultrasound. We have developed an algorithm to warp Baldi into new shapes, and controlling these faces in animated speech. Finally, we are adding resolution to Baldi's wireframe to allow more precise control and more realistic animation. These development efforts and their evaluation will be described.
Bibliographic reference. Cohen, Michael M. / Clark, Rashid / Massaro, Dominic W. (2001): "Animated speech: research progress and applications", In AVSP-2001, 200 (abstract).