The output only resembles 80% of the original. #134
Thanh1122kkll
started this conversation in
General
Replies: 2 comments 1 reply
-
How many photos are you using to create the checkpoint? Gotta tell you did using like 15 photos and the result are amazing. |
Beta Was this translation helpful? Give feedback.
1 reply
-
See my anwser here : #135 |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I have tried various methods, including selecting the full option of the advanced swap mode (codeformer, ldsr, color correction, etc.), and even excluding the blend-face option. However, the output face still has some differences compared to the original. It only approximates about 80% of the original. I wonder if there is a way to achieve a 100% identical result to the original, just swapping then blending the face layer with new skin( like a filter). without adding any additional interpretation. I used a very simple prompt combined with 3 control nets, including lineart, to maximize the level of detail. thanks all
Beta Was this translation helpful? Give feedback.
All reactions