github/DeepFaceLab
1
0
Fork 0
mirror of https://github.com/iperov/DeepFaceLab.git synced 2025-07-06 21:12:07 -07:00
Code Issues Projects Releases Packages Wiki Activity Actions
852 commits 3 branches 2 tags 866 MiB
Commit graph

852 commits

This branch
This branch
All branches
Author SHA1 Message Date
Colombo
59be114485 upd leras 2020-02-27 16:20:14 +04:00
Colombo
acb0b34811 revert 2020-02-27 12:03:01 +04:00
Colombo
30c93a9bdb fix 2020-02-27 11:41:06 +04:00
Colombo
8ad1481209 _ 2020-02-27 11:39:29 +04:00
Colombo
c7ab9653c5 new optimized training: 
for every batch_size*16 samples,
model collects the samples with the highest error and learns them again
therefore hard samples will be trained more often
 2020-02-27 11:19:53 +04:00
Colombo
a5783df546 fix autobackup_hour 2020-02-27 11:13:45 +04:00
Colombo
9860a38907 upd SampleGenerator 2020-02-27 09:58:46 +04:00
Colombo
1898bd6881 _ 2020-02-26 23:50:42 +04:00
Colombo
a6f11cf36b fix 2020-02-26 17:50:36 +04:00
Colombo
f07bbd5fb0 update leras 2020-02-26 13:32:32 +04:00
Colombo
a1cc297cef SAEHD: fix hd arhis 2020-02-25 21:58:05 +04:00
Colombo
b3b97b9802 fix RuntimeWarning during the extraction process 2020-02-25 09:59:48 +04:00
Colombo
28b26ca96c upd README.md 2020-02-25 09:50:31 +04:00
Colombo
0324592b67 update leras/layers.py 2020-02-24 22:23:08 +04:00
Colombo
d65ff888cd fix sort by best. 2020-02-24 20:52:02 +04:00
Colombo
58deb2bed1 enable pretrain option for whole_face type 2020-02-23 20:49:30 +04:00
Colombo
4ea70864fc fix Sorter.py 2020-02-23 20:19:13 +04:00
Colombo
ed384f52f0 upd 2020-02-23 14:21:54 +04:00
Colombo
0a40d8e5da _ 2020-02-22 13:45:00 +04:00
Colombo
75a1adc7c4 fix sort by yaw/pitch for whole_face 2020-02-22 13:43:50 +04:00
Colombo
f1d115b63b added experimental face type 'whole_face' 
Basic usage instruction: https://i.imgur.com/w7LkId2.jpg

	'whole_face' requires skill in Adobe After Effects.

	For using whole_face you have to extract whole_face's by using
	4) data_src extract whole_face
	and
	5) data_dst extract whole_face
	Images will be extracted in 512 resolution, so they can be used for regular full_face's and half_face's.

	'whole_face' covers whole area of face include forehead in training square,
	but training mask is still 'full_face'
	therefore it requires manual final masking and composing in Adobe After Effects.

added option 'masked_training'
	This option is available only for 'whole_face' type.
	Default is ON.
	Masked training clips training area to full_face mask,
	thus network will train the faces properly.
	When the face is trained enough, disable this option to train all area of the frame.
	Merge with 'raw-rgb' mode, then use Adobe After Effects to manually mask, tune color, and compose whole face include forehead.
 2020-02-21 16:21:04 +04:00
Colombo
778fb94246 saehd change face_style help 2020-02-20 20:56:02 +04:00
Colombo
090d4a5be3 fix FANExtractor 2020-02-20 12:20:18 +04:00
Colombo
e6e11ca056 SampleProcessor : added unused random rgb levels 2020-02-20 08:30:53 +04:00
Colombo
c052ee7d60 SAEHD : change help message for eyes_prio : Helps to fix eye problems during training like "alien eyes" and wrong eyes direction ( especially on HD architectures ) by forcing the neural network to train eyes with higher priority. before/after 2020-02-20 08:29:37 +04:00
Colombo
20f842a34a Merger no more terminate process after 60 secs 2020-02-19 22:11:28 +04:00
Colombo
7a790489f2 fix SubprocessorBase.py 2020-02-19 22:09:45 +04:00
Colombo
7a9cd6379f bug fixes 2020-02-19 15:27:22 +04:00
Colombo
144ecd521a MergeMasked fix interpolation 2020-02-19 09:29:27 +04:00
Colombo
712c03b144 _ 2020-02-19 09:15:41 +04:00
Colombo
5d5718704d fix SampleProcessor 2020-02-19 07:00:46 +04:00
Colombo
6aa585fa33 fix leras 2020-02-19 07:00:29 +04:00
Colombo
5ce6f7ef45 Merger: fixed some color flickering in overlay+rct mode 2020-02-19 07:00:14 +04:00
Colombo
f4b163d721 fix 2020-02-18 19:34:23 +04:00
Colombo
febc96c334 upd README.md 2020-02-18 19:16:59 +04:00
Colombo
d4335b5fa5 fix 2020-02-18 19:16:44 +04:00
Colombo
6a8426dcdb upd logo 2020-02-18 19:01:44 +04:00
Colombo
9598ba0141 SAEHD: 
added option Eyes priority (y/n)

	fix eye problems during training  ( especially on HD architectures )
	by forcing the neural network to train eyes with higher priority
	before/after https://i.imgur.com/YQHOuSR.jpg

	It does not guarantee the right eye direction.
 2020-02-18 14:30:07 +04:00
Colombo
4f928074b9 removing smooth_rect option 2020-02-18 10:28:01 +04:00
Colombo
01376fd17c decreased time of training initialization 2020-02-18 10:01:33 +04:00
Colombo
814da70577 Merger: 
added smooth_rect option
	default is ON.
	Decreases jitter of predicting rect by using temporal interpolation.
	You can disable this option if you have problems with dynamic scenes.
 2020-02-17 18:27:09 +04:00
Colombo
e0a55ff1c3 update leras 2020-02-17 18:26:19 +04:00
Colombo
e1635ff760 added CentOS Linux fork 2020-02-15 16:39:56 +04:00
Colombo
60cc917350 add eye masking code 2020-02-03 06:38:58 +04:00
Colombo
4850ef4965 autobackup hour 2020-02-03 06:37:54 +04:00
Colombo
5620763ccf "Enable autobackup" option is replaced by 
"Autobackup every N hour" 0..24 (default 0 disabled), Autobackup model files with preview every N hour
 2020-02-02 20:53:18 +04:00
Colombo
ee27bb0aac Merger: fix bug 2020-02-02 14:45:48 +04:00
Colombo
85792d232a Merger: 
'show alpha mask' now on 'V' button

'super resolution mode' is replaced by
'super resolution power' (0..100) which can be modified via 'T' 'G' buttons
 2020-02-02 10:13:23 +04:00
Colombo
054a7e0e6e Merger: now uses all available CPU cores ( before max 6 ) 
so the more processors, the faster the process will be.
 2020-02-01 18:58:10 +04:00
Colombo
9c07983c68 Merger: 
default erode/blur values are 0.

new multiple faces detection log: https://i.imgur.com/0XObjsB.jpg
 2020-02-01 18:50:43 +04:00