This is pretty cool... the presenters have developed a way to enhance videos both "spatially" (resolution) and "temporally" (slowing down), by using a hybrid combination of low-res video frames with hi-res stills shot from a nearby (as in directly next to it) digital camera at a much lower frame rate... the result is that they are able to warp and map the hi-res detail into the video. Essentially, you could take a standard definition video, and turn it into High Definition video simply by integrating a few interspersed frames of high resolution frames.

Ah... just watch the video: