Newbie question, artefacts on converting

Got questions or tips about the Conversion process? This is the place to discuss them.

Forum rules

Read the FAQs and search the forum before posting a new topic.

This forum is for discussing tips and understanding the process involved with Converting faces from your trained model.

If you are having issues with the Convert process not working as you would expect, then you should post in the Convert Support forum.

Please mark any answers that fixed your problems so others can find the solutions.

Post Reply
User avatar
Posts: 1
Joined: Sat Dec 18, 2021 5:09 pm

Newbie question, artefacts on converting

Post by Zaroff »


I'm pretty new to all this, and inevitably I'm doing something wrong but don't know what.
I've trained a model (Dfl-sae) which is currently around 60k iterations, batch 8. A and B faces are respectively around 4k and 6k. Alignements for the target video were made with both Vgg-obstructed and Bisenet-FP. Video is generated with Vgg.
But, no matter what I do get the artefacts you see in the example screenshot in the link, at the top of the face only.
Does that mean the model isn't trained enough yet, or I made a blunder somewhere else?
Thanks for any helps.


User avatar
Site Admin
Posts: 683
Joined: Fri Jul 12, 2019 12:49 am
Answers: 42
Location: San Francisco
Has thanked: 3 times
Been thanked: 165 times

Re: Newbie question, artefacts on converting

Post by bryanlyon »

The mask you're using isn't removing the extra hair, I suggest using BiSeNet with the hair disabled to remove that.

Post Reply