Multiple extract source for training dataset, how to set alignment file?
Posted: Sun Jan 31, 2021 9:45 am
Hi, I got multiple files to extract training image, some are from pictures and some are from videos (used to swap the extracted face to target video, target video's alignment file is fine...). I noticed that we need alignment file to feed the dataset into the training process in this updated version of software. How to set up the alignment file for it?
Thanks a lot.