Eulerian-Video-Magnification

Eulerian-Video-Magnification copied to clipboard

███████╗██╗ ██╗███╗ ███╗ ██╔════╝██║ ██║████╗ ████║ █████╗ ██║ ██║██╔████╔██║ ██╔══╝ ╚██╗ ██╔╝██║╚██╔╝██║ ███████╗ ╚████╔╝ ██║ ╚═╝ ██║ ╚══════╝ ╚═══╝ ╚═╝ ╚═╝

DESCRIPTION Eulerian Video Magnification for colors and motions magnification

USAGE python evm.py [-h] --video_path VIDEO_PATH [--level LEVEL] [--alpha ALPHA] [--lambda_cutoff LAMBDA_CUTOFF] [--low_omega LOW_OMEGA] [--high_omega HIGH_OMEGA] --saving_path SAVING_PATH [--mode {gaussian,laplacian}] [--attenuation ATTENUATION]

arguments:
  --video_path VIDEO_PATH, -v VIDEO_PATH
                        Path to the video to be used
  --level LEVEL, -l LEVEL
                        Number of level of the Gaussian/Laplacian Pyramid
  --alpha ALPHA, -a ALPHA
                        Amplification factor
  --lambda_cutoff LAMBDA_CUTOFF, -lc LAMBDA_CUTOFF
                        λ cutoff for Laplacian EVM
  --low_omega LOW_OMEGA, -lo LOW_OMEGA
                        Minimum allowed frequency
  --high_omega HIGH_OMEGA, -ho HIGH_OMEGA
                        Maximum allowed frequency
  --saving_path SAVING_PATH, -s SAVING_PATH
                        Saving path of the magnified video (.avi extension necessary)
  --mode {gaussian,laplacian}, -m {gaussian,laplacian}
                        Type of pyramids to use (gaussian or laplacian)
  --attenuation ATTENUATION, -at ATTENUATION
                        Attenuation factor for I and Q channel post filtering

REFERENCES Eulerian Video Magnification for Revealing Subtle Changes in the World (https://people.csail.mit.edu/mrub/evm/)

CONTRIBUTORS Hussem Ben Belgacem