https://github.com/jafagervik/ganban

Pattern Recognition Group 9 Project
https://github.com/jafagervik/ganban

Last synced: 26 days ago
JSON representation

Pattern Recognition Group 9 Project

• Host: GitHub
• URL: https://github.com/jafagervik/ganban
• Owner: Jafagervik
• License: mit
• Created: 2023-05-26T01:46:21.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-07-16T22:54:53.000Z (over 1 year ago)
• Last Synced: 2024-10-30T01:53:58.640Z (2 months ago)
• Language: Python
• Size: 446 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE-MIT

Awesome Lists containing this project

README

        
# GANBan

[![.github/workflows/ci.yml](https://github.com/github/gh-actions-importer/actions/workflows/ci.yml/badge.svg)](https://github.com/github/gh-actions-importer/actions/workflows/ci.yml)

Correcting and coloring old and damaged images using a CycleGAN

| ![image](examples/portrait0_in.png) | ![image](examples/portrait1_in.png) | ![image](examples/portrait2_in.png) | ![image](examples/portrait3_in.png) |

| --- | --- | --- | --- |

| ![image](examples/portrait0_out.png) | ![image](examples/portrait1_out.png) | ![image](examples/portrait2_out.png) | ![image](examples/portrait3_out.png) |

## Converting Images

```sh

python convert.py input_file

```

This will fix a damaged image and and will colorize grayscale images.

This requires the weights `ganban.pth` to be in the same directory. This can be downloaded from [here](https://0x0.st/Hj3_.zip).

Additional arguments are as follows:

```

usage: convert.py [-h] [-i] [-m] [-r] input_file [input_file ...]

Converts an old damaged image into a new colored image. Outputs a file(s) with the name input_file..png

positional arguments:

  input_file          input file(s)

options:

  -h, --help          show this help message and exit

  -i, --intermediate  Output (save) the intermediate states

  -m, --multi-domain  Use the multiple domain/model approach (damaged<>fixed<>color)

  -r, --reverse       Reverses the process (corrected image --> old image)

```

## Dataset

You can download the dataset used [here](https://0x0.st/Hj39.zip). It contains ~900 unique images trained on for 3 different models.

## Structure

### data 

Dataset location and directory to save models in.

### config.py

Contains the hyperparameters.

### cyclegan.py

Contains the generator, discriminator, and initialization weights.

### datasetup.py

Contains the class (Dataset) that interacts with the data

### engine.py

Setup for the train and test steps

### train.py

Entry point to begin training the model. To start, execute `python train.py`.