Skip to content

A reimplementation of the original CycleGAN for image-to-image translation.

Notifications You must be signed in to change notification settings

aanaseer/cyclegan

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CycleGAN

This repository contains a re-implementation of the CycleGAN architecture found in [1].

Experiment tracking was done using Weights and Biases (wandb), selected runs are made public here.

Example

Generating a Van Gogh painting from a photograph and converting it back to a photograph.

Installation and Running

  1. Download the zip file containing preconfigured directories and datasets from here.
  2. Extract the zip file (and all other zip files within) and place contents inside the root directory.
  3. Set up a virtual environment and install the dependencies using the requirements.txt file by running
    pip install -r requirements.txt
    
  4. Amend the configurations dictionary as required in the file train.py.
  5. Train a model by choosing a relevant GPU device with the following command
    CUDA_VISIBLE_DEVICES=[GPU device] python src/train.py
    
    To run without a GPU use,
    python src/train.py
    
  6. Run evaluations by adjusting the configurations dictionary in evaluate.py according to the required settings and by running,
    python src/evaluate.py
    
  7. To compute metrics specify the desired dataset in metrics.py file and then run,
    python src/metrics.py
    
    Currently metrics can only be computed on a CUDA enabled GPU.

Pre-trained Models

Pre-trained models can be downloaded from here.

Notes

  • Hyperparameter tuning was not done for the model.

About

A reimplementation of the original CycleGAN for image-to-image translation.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages