This is a lot like the idea of turning videos / many still images into 3D environments by scary processing stuff and piles of vertices, but this one here actually has some possibility to work, and a much greater usefulness.
Really cool stuff. It would be interesting, if they got this into a full product, to see how high the success rate is. What are the chances that when you zoom in on something you will end up with the right image? It would need a pretty smart indexer, but it would be incredible to see it work. This could change everything and make that graphics hardware useful for something productive!
*Gasp* They seem to have a running preview. (And what's more? It works in Firefox!).
It is a known fact that MS's labs projects are cooler than their big / popular things. The same happens with Google, too.
Really cool stuff. It would be interesting, if they got this into a full product, to see how high the success rate is. What are the chances that when you zoom in on something you will end up with the right image? It would need a pretty smart indexer, but it would be incredible to see it work. This could change everything and make that graphics hardware useful for something productive!
*Gasp* They seem to have a running preview. (And what's more? It works in Firefox!).
It is a known fact that MS's labs projects are cooler than their big / popular things. The same happens with Google, too.