First task is to extract the data required, at least the image components, sound will come later. The augmented data processing with dlib will take perhaps a few days or so, with roughly 1,000,000 images generated from the video files for training and test datasets. I used a GCP lowContinue Reading