Do you know what would be cool? Get the system to recognise heads, hands and bodies the way it does in the games, tell it to flip the projected shape it has recognised and fit it to the other side to create a full 3-d form, and have a system that runs content-aware fills on the shape's missing parts. Do the same for…