Wicky_42 said:
Rig it up to remember the shapes it sees, then you could construct complete 3D models by imaging them from real life with the camera - there's no reason, say, why it couldn't remember what his face looks like from the front when he turns away, or if he turned the other way what the other side looked like.
Pretty sweet demonstration though, shows what the tech is capable of, and also what the limitations of the resolution are - notice how much the background was spiking? I'm guessing you'd need to average out all the noise from each individual point to make a smoother imagine, but you'd lose detail doing that...
yeah that would be great, you turn around a few times and presto you have a fully rendered 3d image of yourself that moves at the same time you do.
As for that matter if it could remember where it is in relation to the room its in you could use it a bit like a scanner and have a 3d image of your entire room.
after that you could get it to recognize what the objects are in the room.
Finally it could then use the available space and objects in a room convert them to look like other objects of similar shape on screen and create a game around it tailored to your room.
edit: to go one step further you could then get your representation to wear different clothes so that you can for instance look like a space marine or pimp