github/DeepFaceLab
1
0
Fork 0
mirror of https://github.com/iperov/DeepFaceLab.git synced 2025-08-20 21:43:21 -07:00
Code Issues Projects Releases Packages Wiki Activity Actions

added new SAE model - Styled AutoEncoder

Browse source
This commit is contained in:
iperov 2019-01-05 09:57:49 +04:00
commit 8e1d337e00
5 changed files with 321 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

13
README.md
View file

@ -70,14 +70,23 @@ LIAEF128 Cage video:
[![Watch the video](https://img.youtube.com/vi/mRsexePEVco/0.jpg)](https://www.youtube.com/watch?v=mRsexePEVco)
- **UFM (2GB+)** - U-net Face Morpher model. If "match_style" option choosed, then this model tries to morph src face to target face and fill around face same background. UFM is result of combining modified U-Net, classic face autoencoder, DSSIM and style losses.
- **UFM (2GB+)** - U-net Face Morpher model with my new face style loss. If "match_style" option choosed, then this model tries to morph src face to target face and fill around face same background. UFM is result of combining modified U-Net, classic face autoencoder, DSSIM and style losses.
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/UFM_Cage_0.jpg)
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/UFM_Cage_1.jpg)
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/DeepFaceLab_convertor_overview.png)
- **SAE (4GB+)** - Styled AutoEncoder. It is like LIAEF but with new face style loss like in UFM.
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/SAE_Cage_0.jpg)
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/SAE_Cage_1.jpg)
- **SAE vs UFM** .
[![Watch the video](https://img.youtube.com/vi/ywiv0_PTp1w/0.jpg)](https://www.youtube.com/watch?v=ywiv0_PTp1w)
![](https://github.com/iperov/DeepFaceLab/blob/master/doc/DeepFaceLab_convertor_overview.png)
### **Sort tool**: