mirror of
https://github.com/Deepshift/DeepCreamPy.git
synced 2025-04-01 10:20:07 +00:00
guide for readme
This commit is contained in:
parent
43d8833d9e
commit
a9393e2acb
@ -34,10 +34,14 @@ In particular, if a vagina or penis is completely censored out, inpainting will
|
||||
# Model
|
||||
The pretrained model can be downloaded from https://drive.google.com/open?id=1mWHYSj0LDSbJQQxjR4hUMykQkVve2U3Q.
|
||||
|
||||
Unzip the contents into the /models/ folder.
|
||||
|
||||
# Usage
|
||||
|
||||
## I. Decensoring hentai
|
||||
|
||||

|
||||
|
||||
The decensorship process is fairly involved. A user interface will eventually be released to streamline the process.
|
||||
|
||||
Using image editing software like Photoshop or GIMP, paint the areas you want to decensor the color with RGB values of (0,255,0). For each censored region, crop 128 x 128 size images containing the censored regions from your images and save them as new ".png" images.
|
||||
@ -67,6 +71,7 @@ $ python train.py
|
||||
|
||||
# To do
|
||||
- Add Python 3 compatibility
|
||||
- Retrain for arbitrary shape censors
|
||||
- Add a user interface
|
||||
- Incorporate GAN loss into training
|
||||
|
||||
|
BIN
readme_images/guide.png
Normal file
BIN
readme_images/guide.png
Normal file
Binary file not shown.
After Width: | Height: | Size: 239 KiB |
Loading…
x
Reference in New Issue
Block a user