MIME-Version: 1.0
Server: CERN/3.0
Date: Sunday, 01-Dec-96 19:52:13 GMT
Content-Type: text/html
Content-Length: 1655
Last-Modified: Friday, 09-Feb-96 04:11:12 GMT
Senior Honors Project at UMass
Design and Implementation of Dynamically
Generated Synchronized Computer Speech And Facial Animation
- Abstract:
- By integrating visual and audio information using computers,
a new form of human-computer interface has been created. The current
research is in the design and implementation of a user interface
whereby a computer can produce speech (audio) and a moving image
of a human face synchronized to the speech (video). The result
is the appearance of a human talking to the user. The program
is given text to speak (for proof-of-concept purposes, the numbers
"0" - "9"); it then creates the speech, automatically
determines the boundaries between phonemes (to synchronize the
video), and uses morphing to smoothly transition between the images
used for each phoneme.
A PostScript copy of the paper
- Sample Output (in Video For Windows .AVI format):
- "Zero"
"One"
"Two"
"Three"
"Four"
"Five"
"Six"
"Seven"
"Eight"
"Nine"
"Hello Professor Hill"