DeepCreamPy/README.md

159 lines
7.5 KiB
Markdown
Raw Normal View History

2018-10-20 04:30:35 +00:00
# DeepCreamPy
2018-10-20 04:00:04 +00:00
*Decensoring Hentai with Deep Neural Networks*
2018-10-20 04:43:56 +00:00
2018-10-20 04:30:35 +00:00
*Formerly named DeepMindBreak*
2018-10-20 04:00:04 +00:00
2018-10-20 04:30:35 +00:00
This project applies an implementation of [Image Inpainting for Irregular Holes Using Partial Convolutions](https://arxiv.org/abs/1804.07723) to the problem of hentai decensorship. Using a deep fully convolutional neural network, DeepCreamPy can replace censored artwork in hentai with plausible reconstructions. The user needs to only specify the censored regions.
2018-10-20 04:00:04 +00:00
![Censored, decensored](/readme_images/mermaid_collage.png)
# What's New?
- Decensoring images of ANY size
- Higher quality decensors
2018-10-21 05:37:43 +00:00
- Support for mosaic decensors
2018-10-20 04:00:04 +00:00
- User interface (still a WIP and not very usable)
# Limitations
2018-10-21 05:39:03 +00:00
The decensorship is intended to ONLY work on color hentai images that have minor to moderate bar/mosaic censorship of the penis or vagina.
2018-10-20 22:12:38 +00:00
2018-10-20 04:00:04 +00:00
It does NOT work with:
- Black and white images
- Monochrome images
- Hentai containing screentones (e.g. printed hentai)
- Real life porn
- Censorship of nipples
- Censorship of anus
- Animated gifs/videos
In particular, if a vagina or penis is completely censored out, decensoring will be ineffective.
2018-10-20 22:51:03 +00:00
# Releases
You can download the latest release [here](https://github.com/deeppomf/DeepCreamPy/releases/latest) or find all previous releases [here](https://github.com/deeppomf/DeepCreamPy/releases).
If you want to run the code yourself, you can download the model from https://drive.google.com/open?id=1byrmn6wp0r27lSXcT9MC4j-RQ2R04P1Z. Unzip the file into the /models/ folder.
2018-10-20 04:00:04 +00:00
2018-10-20 22:51:03 +00:00
# Dependencies
2018-10-20 04:00:04 +00:00
- Python 3
- TensorFlow 1.10
- Pillow
2018-10-20 07:35:45 +00:00
- h5py
2018-10-20 04:00:04 +00:00
No GPU required! Tested on Ubuntu 16.04 and Windows. (Tensorflow on Windows is compatible with Python 3 and not Python 2.)
2018-10-20 22:51:03 +00:00
Tensorflow, Pillow, and h5py can all be installed by running in the command line
2018-10-20 22:12:38 +00:00
```
$ pip install -r requirements.txt
```
2018-10-20 04:00:04 +00:00
# Usage
2018-10-20 04:42:48 +00:00
## I. Decensoring bar censors
2018-10-20 06:10:26 +00:00
For each image you want to decensor, using image editing software like Photoshop or GIMP to color the areas you want to decensor the green color (0,255,0), which is a very bright green color.
2018-10-20 04:00:04 +00:00
2018-10-20 06:10:26 +00:00
*I strongly recommend you use the pencil tool and NOT the brush tool.*
*If you aren't using the pencil tool, BE SURE TO TURN OFF ANTI-ALIASING on the tool you are using.*
2018-10-20 04:00:04 +00:00
I personally use the wand selection tool with anti-aliasing turned off to select the censored regions. I then expand the selections slightly, pick the color (0,255,0), and use the paint bucket tool on the selection regions.
To expand selections in Photoshop, do Selection > Modify > Expand or Contract.
To expand selections in GIMP, do Select > Grow.
2018-10-20 05:27:53 +00:00
Save these images in the PNG format to the "decensor_input" folder.
### A. Using the binary
2018-10-20 05:27:53 +00:00
2018-10-20 05:30:20 +00:00
Decensor the images by double-clicking on the decensor file.
2018-10-20 05:27:53 +00:00
2018-10-20 05:30:20 +00:00
### B. Running from scratch
2018-10-20 05:27:53 +00:00
Decensor the images by running
2018-10-20 04:00:04 +00:00
```
$ python decensor.py
```
2018-10-20 07:17:33 +00:00
Decensored images will be saved to the "decensor_output" folder. Decensoring takes a few minutes per image.
2018-10-20 04:00:04 +00:00
2018-10-20 04:42:48 +00:00
## II. Decensoring mosaic censors
As with decensoring bar censors, perform the same steps of coloring the censored regions green and putting the colored image into the "decensor_input" folder.
In addition, move the original images (before you colored them) into the "decensor_input_original" folder. Ensure each original image has the same names as their corresponding colored version in the "decensor_input" folder.
For example, if the original image is called "mermaid.jpg," then you want to put this image in the "decensor_input_original" folder and, after you colored the censored regions, name the colored image "mermaid.png" and move it to the "decensor_input" folder.
### A. Using the binary
2018-10-21 05:37:43 +00:00
Decensor the images by double-clicking on the decensor_mosaic file.
### B. Running from scratch
Decensor the images by running
```
$ python decensor.py --is_mosaic=True
```
Decensored images will be saved to the "decensor_output" folder. Decensoring takes a few minutes per image.
2018-10-20 04:42:48 +00:00
2018-10-21 05:37:43 +00:00
## III. Decensoring with the user interface
2018-10-20 04:42:48 +00:00
To be implemented.
2018-10-20 07:17:33 +00:00
# Troubleshooting
If your decensor output looks like this:
2018-10-20 07:18:30 +00:00
2018-10-20 07:17:33 +00:00
![Bad decensor](/readme_images/mermaid_face_censored_bad_decensor.png)
2018-10-20 07:18:30 +00:00
2018-10-20 07:18:57 +00:00
then the censored regions were not colored correctly.
2018-10-20 07:17:33 +00:00
2018-10-20 07:18:57 +00:00
*Make sure you have antialiasing off.*
2018-10-20 07:17:33 +00:00
Here are some examples of bad and good colorings:
2018-10-20 07:23:26 +00:00
|Image|Zoom|Comment|
|--- | --- | ---|
|![Incomplete coloring](/readme_images/mermaid_face_censored_bad_incomplete.png)|![Incomplete coloring](/readme_images/mermaid_face_censored_bad_incomplete_zoom.png)|Some censored pixels was left uncolored. Expand your selections to fully cover all censored regions.|
|![Bad edges](/readme_images/mermaid_face_censored_bad_edge.png)|![Bad edges](/readme_images/mermaid_face_censored_bad_edge_zoom.png)|Some pixels around the edges of the green regions are not pure green. This will cause the green to bleed into the decensors. Make sure anti-aliasing is off and to use a pencil tool and not a brush tool if possible.|
2018-10-21 05:39:03 +00:00
|![Perfect coloring!](/readme_images/mermaid_face_censored_good.png)|![Perfect coloring! The censored region is uniformly colored correctly.](/readme_images/mermaid_face_censored_good_zoom.png)|Perfect coloring!|
2018-10-20 07:17:33 +00:00
2018-10-20 04:00:04 +00:00
# To do
- Finish the user interface
Contributions are welcome! Special thanks to StartleStars for contributing code for mosaic decensorship and SoftArmpit for greatly simplifying decensoring!
# License
2018-10-20 04:42:48 +00:00
**I will likely change the license to be more permissive in the future, but I make no guarantees.**
2018-10-20 04:43:56 +00:00
2018-10-20 04:44:15 +00:00
**This work (the code and the model) is under my exclusive copyright. The only right I grant to users is using this work for personal use to decensor hentai. Copying, distribution, and modification is prohibited. Commercial use is prohibited. (Modification of the 6 values of the arguments in config.py is allowed.)**
2018-10-20 04:00:04 +00:00
Example image by Shurajo & AVALANCHE Game Studio under [CC BY 3.0 License](https://creativecommons.org/licenses/by/3.0/). The example image is modified from the original, which can be found [here](https://opengameart.org/content/mermaid).
Neural network code is modified from MathiasGruber's project [Partial Convolutions for Image Inpainting using Keras](https://github.com/MathiasGruber/PConv-Keras), which is an unofficial implementation of the paper [Image Inpainting for Irregular Holes Using Partial Convolutions](https://arxiv.org/abs/1804.07723).
User interface code is modified from Packt's project [Tkinter GUI Application Development Blueprints - Second Edition](https://github.com/PacktPublishing/Tkinter-GUI-Application-Development-Blueprints-Second-Edition).
```
# Copyright (c) 2018 MathiasGruber, Packt
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in all
# copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
```