The challenge with adapting any algorithm that operates on photos to video is ensuring consistency across frames so the effect isn't jarring. That said, perhaps

The challenge with adapting any algorithm that operates on photos to video is ensuring consistency across frames so the effect isn't jarring. That said, perhaps not surprisingly, there has been work in solving this which seems to have started with the paper below. http://www.faculty.idc.ac.il/arik/SCWeb/vidret/index.html http://www.faculty.idc.ac.il/arik/SCWeb/vidret/index.html