Holy crap. That’s probably the first thing that you’ll utter when you watch the Fotowoosh demo video. A 3D scene from a single 2D photo? If this isn’t vaporware, it’s going to be a hell of a lot of fun.
Nothing new… it’s basically MetaCreation’s Canoma from 1999.
I remember using Canoma — very cool. But this is meant to be pretty much automated on the other hand. Effectively recognizing perspective from the picture (vs. mapping objects onto it). This is exciting as you can imagine it happening in real-time, in a slideshow for example. Of course, we won’t know how well this works until the final application is out.
It seems to be the company built on this CMU research