Via Wikinomics and TechEBlog, Microsoft has more interesting tools in development: Photo Tourism assembles navigable 3d models from photos (U of Wasington research page) and Unwrap Mosaic (Microsoft page) can Reconstruct 3D Surface Models From Video.
Also on New Scientist
Update: Courtesy of Boing-Boing and Engadget, you might see another round of posts on Using Photographs to Enhance Videos of a Static Scene (video below). Mark Christiansen posted on this about a year ago with interesting details on the Adobe connection. The movies were mentioned on the web numerous times after that (after SIGGRAPH) last year, including here in Enhanced photogrammetry and 2007 image processing projects.
As noted here earlier in Multidimensional photos, Blaise Aguera y Arcas demonstrated Microsoft's Seadragon and Photosynth at the TED in 2007.
Update 2: Adobe research was discussed here recently in Adobe Advanced Technology: more than limited color picker.
Update 3: Indie Mogul adds value with a few more movies explaining recent papers on graphics research:
Interactive Video Cutout from pro on Vimeo.
GradientShop: A novel approach to image and video processing from pro on Vimeo.