Ok, this is amazing. And nerdy. Students at the University of Washington have developed a system to improve video with still photographs. If I understand correctly, they have written software that uses video of a static scene footage to produce a 3D model of the scene. Then they can apply the higher resolution of a photograph taken of the same scene back onto the video. They also do some other awesome stuff. The most amazing thing is probably where they remove a no parking sign from a shot. They replace the background that the sign is covering up by using information from other frames of the shot. Wow.