“Hyperlapse” processing—which analyzes a video, reconstructs the 3D input camera path, and then synthesizes a smooth new camera path—could pair brilliantly with Google Glass, GoPro cameras, and other wearable capture devices:
Here’s a more technical description of how it works:
WOW! That looks awesome. How soon before its available!?!
Breathtaking and inspired. But my first thought was will someone pair this with Google Streetview and let you mark 2 points on a map and generate one of these smooth speedrun videos?