[jcoffland]

I wonder of you could use this on car commercials and thereby deduce the actual comparative quality of construction. Assuming that better built cars would vibrate less at speed. It would be interesting to understand where the vibrations occur.

A question for the authors. All the video is with vibrating but otherwise stationary objects. Must the object be stationary for these tools to work?

[initram]

Probably not because in a lot of car videos there's no real car. Here's one simple example: https://www.youtube.com/watch?v=b7vTM4_rjhs

There's a company that makes a rig that's basically just an engine and 4 wheels. You film your entire commercial using it and they replace it with computer imagery of your car in post. I can't find the video at the moment, though.

[yorwba]

Not one of the authors, but I had a look at their paper http://www.interactivedynamicvideo.com/ISMB_Davis_2015.pdf

They use optical flow to get displacement values for each pixel. This works well when you don't have very large movements between frames, no sudden illumination changes and no rotations.

Car commercials will probably have all of those and on top of that be highly edited (e.g. you have to deal with cuts in the video)

You could try to use some other approach for motion estimation, like identifying and tracking salient points, but that would require quite a lot of work on top of what this demo shows.

[infinite8s]

Not the author, but I surmise it would be much more difficult to assign vibration modes to subparts of the image in the image without being able to track them across frames.