Skip to content

A Tensorflow implementation of CycleGAN using Eager Execution, tf.keras.layers, and tf.data.

License

Notifications You must be signed in to change notification settings

herbiebradley/CycleGAN-Tensorflow

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CycleGAN-Tensorflow

A Tensorflow implementation of Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks using Eager Execution, tf.keras.layers, and tf.data.

Requirements:

  • Tensorflow 1.11

Thanks to the original authors PyTorch implementation for inspiration: https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix

Project Organization

├── README.md
├── requirements.txt   <- Use `pip install -r requirements.txt`
├── setup.py           <- makes project pip installable (pip install -e .) so src can be imported
├── LICENSE
└── src                <- Source code for use in this project
    ├── __init__.py    <- Makes src a Python module
    │
    ├── train.py       <- Run this to train
    │
    ├── test.py        <- Run this to test
    │
    ├── data           <- Code for downloading or loading data  
    │   ├── data.py         <- Dataset class
    │   └── download_data.py
    │
    ├── models         <- Code for defining the network structure and loss functions
    │   ├── cyclegan.py     <- CycleGAN model class
    │   ├── networks.py
    │   └── losses.py
    │
    └── utils          <- Utility files
        ├── options.py      <- Class for command line options
        └── image_history_buffer.py

Project organisation based on the cookiecutter data science project template. #cookiecutterdatascience

About

A Tensorflow implementation of CycleGAN using Eager Execution, tf.keras.layers, and tf.data.

Topics

Resources

License

Stars

Watchers

Forks

Languages