Image editing just got easier. Computer graphics researchers at Carnegie Mellon University have developed systems that could allow people to edit photographs using the millions of images available on the web.
Even the graphically challenged can add people or objects to a photo — or simply fill in holes. The systems automatically find images that match the context of the original photo so they blend realistically. (View a before & after slideshow.)
A photo showing a vacant street, for example, could be populated with people, vehicles and even parking meters. Or a passerby might be erased from a shot of a group of friends.
"We are able to leverage the huge amounts of visual information available on the Internet to find images that make the best fit," said Alexei A. Efros, assistant professor of computer science and robotics. "It's not applicable for all photo editing, such as when an image of a specific object or person is added to a photo. But it's good enough in many cases. Why Photoshop if you can 'photoswap' instead?"
Efros and his colleagues are presenting papers on two systems at the Association for Computing Machinery's Special Interest Group on Graphics and Interactive Techniques (SIGGRAPH) annual conference Aug. 5-9 in San Diego.
One of the systems — called Photo Clip Art — was developed by Carnegie Mellon grad students Jean-François Lalonde and Derek Hoiem with Carsten Rother, John Winn and Antonio Criminisi of Microsoft Research Cambridge. It uses thousands of labeled images from a website called LabelMe as clip art that can be added to photos.
The other system — called Scene Completion — was developed by grad student James Hays, another member of Efros' research team. It draws on millions of photos from Flickr to fill in holes in photos resulting from damage to a photograph or an editor's cuts.
"Matching an object with the original photo and placing that object within the 3-D landscape of the photo is a complex problem," said Lalonde, who led development of the system. "But with our approach, and a lot of clip art data, we can hide the complexity from the user and make the process simple and intuitive."