can you share a bit how did you preprocess the ground truth of the image? #87

Open
opened 2026-01-29 21:41:33 +00:00 by claunia · 0 comments
Owner

Originally created by @jorjiang on GitHub (Oct 19, 2021).

I tried to retrain the model with a different setup. as you mentioned that I might need to preprocess the image with some beauty filter, which is what you also did.

I have applied some simple beauty filters, but the result is not quite as good as yours, for example, the results are a little bit blurry and the eye lashes are much less.

could you share a bit what kind of preprocessing you have done to the ground truth to achieve the current results?

Originally created by @jorjiang on GitHub (Oct 19, 2021). I tried to retrain the model with a different setup. as you mentioned that I might need to preprocess the image with some beauty filter, which is what you also did. I have applied some simple beauty filters, but the result is not quite as good as yours, for example, the results are a little bit blurry and the eye lashes are much less. could you share a bit what kind of preprocessing you have done to the ground truth to achieve the current results?
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: TencentARC/GFPGAN#87