If Oliver Kreylos' Kinect 3D video capture work was amazing already, his hack to use two Kinects to eliminate shadows and fully capture reality in 3D space is just crazy.
Is AI Moving Too Fast? | Future Tech
In Oliver's previous experiment, the 3D video capture was limited by Kinect's single point of view. This point of view created shadows of nonexistent data, since Microsoft's camera can't see behind objects. To solve this problem, he added a second Kinect at a different angle to capture what the other Kinect couldn't see. By combining both video signals and three-dimensional information, he was able to fully recreate three-dimensional reality, with no shadows. [Okreylos YouTube Channel]