How to training / convert >= 2 faces included on same picture?
How to training / convert >= 2 faces included on same picture?
Read the FAQs and search the forum before posting a new topic.
This forum is for discussing tips and understanding the process involved with Training a Faceswap model.
If you have found a bug are having issues with the Training process not working, then you should post in the Training Support forum.
Please mark any answers that fixed your problems so others can find the solutions.
Re: How to training / convert >= 2 faces included on same picture?
Short but sweet method.
- Do normal extraction.
- Copy the alignments file, so you have 2 files (1 for person 1, 1 for person 2).
- Copy the extracted faces folder (so you have 2 identical folders, 1 for person 1, 1 for person 2)
- Do the standard cleaning/sorting for person 1 with alignments file 1 and faces folder 1
- Do the standard cleaning/sorting for person 2 with alignments file 2 and faces folder 2
When you come to convert, run convert with the first alignment file with your first model.
For the 2nd person, use the 2nd alignment file, with the 2nd model, and input the video that was output from your first conversion.
My word is final