VMAF is a perceptual video quality assessment algorithm developed by Netflix. It is based on the idea that the quality of a video is determined by the viewer's perception of the video, rather than the technical specifications of the video itself. VMAF uses a combination of machine learning and computer vision techniques to predict the perceived quality of a video based on a set of features extracted from the video.
For more information, see the VMAF GitHub repository.
You can check if VMAF is installed by running the following command:
ffmpeg -helpIf --enable-libvmaf is in the output, VMAF is installed. Otherwise, you need to install it.
Install latest git build from here, then update the PATH:
wget https://johnvansickle.com/ffmpeg/builds/ffmpeg-git-amd64-static.tar.xz
tar -xvf ffmpeg-git-amd64-static.tar.xz
export PATH=$PATH:/path/to/ffmpeg-git-20220307-amd64-static
Test the installation with:
which ffmpeg
ffmpeg -version
ffmpeg -filters | grep vmaf
It should output the path to the ffmpeg binary, the version of ffmpeg and the vmaf filter.