This video is the supplemental to our SIGGRAPH 2013 paper "Phase-Based Video Motion Processing". See our website for more information: people.csail.mit.edu/nwadhwa/p...
Cool...could you use this to reduce atmospheric noise in telescope observations? Also what about looking at stuff through microscopes and viewing video captures at different frequency ranges, very interesting! And, I think you could find more interesting stuff looking through light fields and running your algorithms in 3d (after calculating 3d from the light field)
you might consider using this technic to create a star trek style medical tricorder... your method is passive... have you considered sending something like a focused sound or light wave at a target and reading the reaction? kinda like tapping something to see what its made of
has there been any research to enhance the quality of the original video using the detected motion data? what I mean is - original footage has these motion frequencies in it, but, when the amplification is performed and highest amplitude motion frequencies are found, wouldn't creating a filter based on these frequencies and applying it to the original video make it sharper? -> in essence - filter out any micro movement which might smear edges
Cool...could you use this to reduce atmospheric noise in telescope observations? Also what about looking at stuff through microscopes and viewing video captures at different frequency ranges, very interesting! And, I think you could find more interesting stuff looking through light fields and running your algorithms in 3d (after calculating 3d from the light field)
Wonderful explanation!! thanks for sharing!
The origin of #MotionAmplification or #MotionMagnification great video folks! Love it.
This is so sick. 10/10.
Excellent work.
you might consider using this technic to create a star trek style medical tricorder... your method is passive... have you considered sending something like a focused sound or light wave at a target and reading the reaction? kinda like tapping something to see what its made of
One more very interesting video that I understand only a fraction of
Very fascinating. What chair is that from? Pattern recognition? Theoretical Computer Science?
Nice video thanks!
great video
why dont you use 1/8th octave pyramids... or 1/16th or 1/32?
amazing
Need this in an executable format asap
Keep Up the Great WORK
has there been any research to enhance the quality of the original video using the detected motion data?
what I mean is - original footage has these motion frequencies in it, but, when the amplification is performed and highest amplitude motion frequencies are found, wouldn't creating a filter based on these frequencies and applying it to the original video make it sharper? -> in essence - filter out any micro movement which might smear edges
great
Does this fail if the camera is hand held? or if the tripod suffers from faint quakes?
It seems like it would amplify camera shake movements too, so it would be REALLY jumpy.
I think it will more or less ignore hand shaking with the frequency setting. The amplification only shows significantly at it's own frequency.
woah
how to collect this in Nuke)?
We need an app hurry please.
is this a plugin for any video software yet?
The papers are public, you can implement it yourself
how do i run the code?
Can someone please tell me what on earth he is talking about, I have no idea.