The software uses source photos to animate a facial expression from a smile to a frown or deliver a visualization how a face ages over time. The basic idea is image morphing, but the project aims for much more sophisticated face-detection software in the same spirit that created research at the same university that resulted in Microsoft's Photosynth 3D photo-stitching software.
"I have 10,000 photos of my 5-year-old son, taken over every possible expression," said Steve Seitz, a UW professor of computer science and engineering and engineer in Google's Seattle office. "I would like to visualize how he changes over time, be able to see all the expressions he makes, be able to see him in 3-D or animate him from the photos." Seitz will be discussing the software this fall at the International Conference on Computer Vision.
A first version of the software has been made available some time ago as the Face Movie feature with Google's Picasa. The Face Movie version includes some simplifications to make it run more quickly, the researchers said. "There's been a lot of interest in the computer vision community in modeling faces, but almost all of the projects focus on specially acquired photos, taken under carefully controlled conditions," Seitz said. "This is one of the first papers to focus on unstructured photo collections, taken under different conditions, of the type that you would find in iPhoto or Facebook."
The researchers believe that the software could deliver much more realistic avatars in future and provide a better bridge from the real to the digital world.