github/DeepFaceLab
1
0
Fork 0
mirror of https://github.com/iperov/DeepFaceLab.git synced 2025-07-06 04:52:13 -07:00
Code Issues Projects Releases Packages Wiki Activity Actions

changed help message for pixel loss:

Browse source 
Pixel loss may help to enhance fine details and stabilize face color. Use it only if quality does not improve over time.

SAE:
previous SAE model will not work with this update.
Greatly decreased chance of model collapse.
Increased model accuracy.
Residual blocks now default and this option has been removed.
Improved 'learn mask'.
Added masked preview (switch by space key)

Converter:
fixed rct/lct in seamless mode
added mask mode (6) learned*FAN-prd*FAN-dst

added mask editor, its created for refining dataset for FANSeg model, and not for production, but you can spend your time and test it in regular fakes with face obstructions
This commit is contained in:
iperov 2019-04-04 10:22:53 +04:00
parent 01e98cde8e
commit 5ac7e5d7f1
22 changed files with 715 additions and 387 deletions
Download patch file Download diff file Expand all files Collapse all files

2
samplelib/SampleProcessor.py
View file

@ -152,7 +152,7 @@ class SampleProcessor(object):
if is_face_sample:
if face_mask_type == 1:
img = np.concatenate( (img, LandmarksProcessor.get_image_hull_mask (img.shape, cur_sample.landmarks) ), -1 )
img = np.concatenate( (img, LandmarksProcessor.get_image_hull_mask (img.shape, cur_sample.landmarks, cur_sample.ie_polys) ), -1 )
elif face_mask_type == 2:
mask = LandmarksProcessor.get_image_eye_mask (img.shape, cur_sample.landmarks)
mask = np.expand_dims (cv2.blur (mask, ( w // 32, w // 32 ) ), -1)