Python module to split videos into raw frames and audio samples and create an HDF5 dataset
Usage:
from videodemux import Demux
images, audio = Demux().run(input_filename)
images is a numpy array containing raw 8-bit RGB video frames. audio is a numpy array of signed 16-bit stereo PCM samples.