what is the difference between legacy preprocessing and preprocessing?
can i use crop_mouth_from_video.py to preprocess part of LRW dataset not all 500 words folders, and how what should i change?
preprocessing output will be stored in which format?
Thank you