Communications of the ACM
Oscar-Winning Animator and Filmmaker Chris Landreth at of T Computer Science
Academy Award-winning animator Chris Landreth is working on the development of software that can predict an animated character's facial expressions from the sound of an actor's voice.
Credit: National Film Board of Canada
In his capacity as distinguished research artist-in-residence at the University of Toronto's computer science department, Academy Award-winning animator and filmmaker Chris Landreth is pursuing the creation of software that can predict an animated character's facial expressions using only the sound of an actor's voice.
Landreth and Ph.D. student Pif Edwards are combining speech detection and animation to seamlessly generate virtual actors, or scenes of crowds in which hundreds of people must talk and emote. The software does this by listening to a voice, picking out the phoneme sounds, producing a streaming list of phonemes, and applying it to an animated character.
Landreth expects the predictive software's applications to extend beyond animation. He notes capturing, displaying, and changing a computer graphic model on demand can be a helpful therapeutic tool with uses ranging from psychological therapy research to helping autistic people.
"This isn't just production," Landreth says. "Or just animation. My experience being involved in facial research is that it spans many disciplines, which is what I find wonderful about it."
From U of T News
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found