github/DeepFaceLab
1
0
Fork 0
mirror of https://github.com/iperov/DeepFaceLab.git synced 2025-07-15 01:23:44 -07:00
Code Issues Projects Releases Packages Wiki Activity Actions
1140 commits 3 branches 2 tags 866 MiB
Commit graph

134 commits

Author SHA1 Message Date
iperov
68e55baa15 added Trump speech as example of deepfake 2021-01-09 14:47:11 +04:00
iperov
637c2b2a9b upd political_speech2.jpg 2020-12-26 10:52:41 +04:00
iperov
fc7ed4d0ee readme: upd faces 2020-12-21 02:01:18 +04:00
iperov
e391b5aa4d update face examples 2020-12-20 23:26:41 +04:00
iperov
b12746fb8a readme: update face examples 2020-12-20 22:28:48 +04:00
Colombo
6903841161 update face examples 2020-10-15 19:28:01 +04:00
Colombo
d91e79fa64 README: added two new politician manipulation videos 2020-10-02 15:12:09 +04:00
Colombo
0efebe6bc5 upd deepfake_progress.png 2020-07-17 15:56:35 +04:00
Colombo
6c7bb74ad4 README: update deepfake_progress, 448 pix 2020-07-16 22:26:43 +04:00
Colombo
e5e9f1689c readme: change second headswap example 2020-07-03 13:11:20 +04:00
Colombo
f0508287f0 added de-age the face example 2020-06-24 10:15:44 +04:00
Colombo
9fd3a9ff8d add meme 2020-06-18 22:15:25 +04:00
Colombo
9994f1512e added third head swap example 2020-06-18 13:56:42 +04:00
Colombo
11327cb0c5 upd face examples 2020-06-17 10:51:11 +04:00
Colombo
70182f8c65 upd face examples 2020-06-17 10:49:41 +04:00
Colombo
2e650a7122 . 2020-06-12 10:45:57 +04:00
Colombo
dc43f5a891 add second headswap example 2020-06-09 16:01:47 +04:00
Colombo
65bc9273e6 upd readme 2020-06-04 15:38:20 +04:00
Colombo
903657ff91 upd readme 2020-06-04 14:39:50 +04:00
Colombo
130d72cb8b upd 2020-05-22 22:44:00 +04:00
Colombo
72e2f771a2 upd readme 2020-05-20 09:58:55 +04:00
Colombo
8ddc7954da upd README.md : added mini tutorial 2020-04-13 20:57:51 +04:00
Colombo
b6b3936bcd update readme 2020-04-06 20:15:46 +04:00
Colombo
453237bfd7 upd README.md 2020-04-06 17:55:33 +04:00
Colombo
4db11aa133 upd README.md 2020-04-04 09:40:18 +04:00
Colombo
dbb0988927 upd README.md 2020-04-01 22:17:59 +04:00
Colombo
3a9f22f68e upd README.md 2020-03-24 16:21:32 +04:00
Colombo
883f72867f Alipay 捐款 2020-03-22 20:56:57 +04:00
Colombo
7015f6fe09 QQ 捐款 2020-03-22 20:39:21 +04:00
Colombo
ed8af226a3 QQ 捐款 2020-03-22 20:37:55 +04:00
Colombo
c6e4ae8613 _ 2020-03-21 14:40:27 +04:00
Colombo
51ec884d96 _ 2020-03-21 14:33:49 +04:00
Colombo
f9684cb155 _ 2020-03-21 14:16:55 +04:00
Colombo
9b11a9c9a4 add QQ 捐款 2020-03-21 10:42:08 +04:00
Colombo
94e38a43fd upd readme 2020-03-15 22:42:24 +04:00
Colombo
f07349cc32 upd logo_python.png 2020-03-09 08:05:04 +04:00
Colombo
d5c29b55e3 upd README.md 2020-03-08 14:48:19 +04:00
Colombo
ed384f52f0 upd 2020-02-23 14:21:54 +04:00
Colombo
6a8426dcdb upd logo 2020-02-18 19:01:44 +04:00
Colombo
430c2f974e upd README.md 2020-01-31 22:32:21 +04:00
Colombo
4ed16c67f5 upd readme 2020-01-31 14:24:03 +04:00
Colombo
38b85108b3 DFL-2.0 initial branch commit 2020-01-21 18:43:39 +04:00
Colombo
a347bb4f15 change doc about builds 2019-12-27 14:29:27 +04:00
Colombo
b93ae56a08 upd readme 2019-12-23 19:15:59 +04:00
Colombo
4868bd1324 upd doc/doc_prebuilt_windows_app.md 2019-11-12 09:16:42 +04:00
Colombo
2ecf603d96 update prebuilt app info page 2019-11-07 15:07:22 +04:00
Colombo
9b08199e87 upd .md 2019-10-09 12:01:37 +04:00
Colombo
ac7725163d removed SAEv2, 
added SAEHD model ( High Definition Styled AutoEncoder )
This is a new heavyweight model for high-end cards to achieve maximum possible deepfake quality in 2020.
Differences from SAE:
+ new encoder produces more stable face and less scale jitter
  before: https://i.imgur.com/4jUcol8.gifv
  after:  https://i.imgur.com/lyiax49.gifv - scale of the face is less changed within frame size
+ new decoder produces subpixel clear result
+ pixel loss and dssim loss are merged together to achieve both training speed and pixel trueness
+ by default networks will be initialized with CA weights, but only after first successful iteration
  therefore you can test network size and batch size before weights initialization process
+ new neural network optimizer consumes less VRAM than before
+ added option <Enable 'true face' training>
  The result face will be more like src and will get extra sharpness.
  example: https://i.imgur.com/ME3A7dI.gifv
  Enable it for last 15-30k iterations before conversion.
+ encoder and decoder dims are merged to one parameter encoder/decoder dims
+ added mid-full face, which covers 30% more area than half face.
 2019-10-08 15:09:28 +04:00
Colombo
9e80bbc917 1 2019-09-21 15:19:58 +04:00
iperov
63d30c49ae fixes and optimizations, 
converters: added new option sharpen_mode and sharpen_amount
 2019-08-25 19:18:51 +04:00