It has been the innovative thinking of D.Hoiem, A.A. Efros, M Herbert from Carnegie Mellon University. The video shows the use of a single image for forming three-dimensional perception. It has been done by managing the computer to learn the geometric context of a 2D image automatically.

Futuristic plans are to use the technology in robots eyes for better understanding of the images. This ability is developed by teaching the computers to spot visual cues to differentiate horizontal and vertical surfaces in outdoor scenes.

Via: mantas 37609