https://github.com/ghaiszaher/foggy-cyclegan

Fog Simulation using Generative Adversarial Networks (GAN). This code is the implementation of the master thesis Simulating Weather Conditions on Digital Images. It uses a modified CycleGAN model to synthesize fog on clear images.
https://github.com/ghaiszaher/foggy-cyclegan

cyclegan cyclegan-tensorflow deep-learning dissertation fog gan generative-adversarial-network image-processing msc-thesis simulating-weather-conditions

Last synced: 4 days ago
JSON representation

Fog Simulation using Generative Adversarial Networks (GAN). This code is the implementation of the master thesis Simulating Weather Conditions on Digital Images. It uses a modified CycleGAN model to synthesize fog on clear images.

• Host: GitHub
• URL: https://github.com/ghaiszaher/foggy-cyclegan
• Owner: ghaiszaher
• License: gpl-3.0
• Created: 2020-03-05T20:19:11.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2024-12-24T00:44:21.000Z (14 days ago)
• Last Synced: 2024-12-27T11:10:05.985Z (11 days ago)
• Topics: cyclegan, cyclegan-tensorflow, deep-learning, dissertation, fog, gan, generative-adversarial-network, image-processing, msc-thesis, simulating-weather-conditions
• Language: Python
• Homepage: http://ghaiszaher.me/Foggy-CycleGAN/
• Size: 67.1 MB
• Stars: 73
• Watchers: 2
• Forks: 13
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
> [!NOTE]  

> November 2024: New Pre-trained Models are available, check the [Pre-trained Models](#pre-trained-models) section.

# Foggy-CycleGAN



 



This project is the implementation for my Computer Science MSc thesis in the University of Debrecen.

Dissertation: 

[PDF] Simulating Weather Conditions on Digital Images (Debrecen, 2020).

# Table of Content

- [Description](#description)

- [Code](#code)

- [Notebook](#notebook-)

- [Results (2020)](#results-2020)

- [Pre-trained Models](#pre-trained-models)

- [Results (2024)](#results-2024)

- [2024-11-17-rev1-000 Test Notebook](#2024-11-17-rev1-000-test-notebook-)

## Description

**Foggy-CycleGAN** is a

A Jupyter Notebook file Foggy_CycleGAN.ipynb is available in the repository.

## Code

The full source code is available under GPL-3.0 License in my Github repository ghaiszaher/Foggy-CycleGAN

## Notebook 

A Jupyter Notebook file Foggy_CycleGAN.ipynb is available in the repository.

## Results (2020)



 





 





 





 





© Ghais Zaher 2020



## Pre-trained Models

As legacy pre-trained models are no longer compatible with newer Keras/Tensorflow versions, I have retrained the model and made the new weights available to download.

Each of the following models was trained in Google Colab using the same dataset, the parameters for building the models and number of trained epochs are a bit different:



  

    

      Model

      Trained Epochs

      Config

    

  

  

    

      2020-06 (legacy)

      145

      

        use_transmission_map=False


        use_gauss_filter=False


        use_resize_conv=False

      

    

    

      2024-11-17-rev1-000

      522

      

        use_transmission_map=False


        use_gauss_filter=False


        use_resize_conv=False

      

    

    

      2024-11-17-rev2-110

      100

      

        use_transmission_map=True


        use_gauss_filter=True


        use_resize_conv=False

      

    

    

      2024-11-17-rev3-111

      103

      

        use_transmission_map=True


        use_gauss_filter=True


        use_resize_conv=True

      

    

    

      2024-11-17-rev4-001

      39

      

        use_transmission_map=False


        use_gauss_filter=False


        use_resize_conv=True

      

    

  



## Results (2024)

The results of the new models are similar to the previous ones, here are some samples:



  

    

      Clear

      2024-11-17-rev1-000

      2024-11-17-rev2-110

      2024-11-17-rev3-111

      2024-11-17-rev4-001

    

  

  

    

      

      

      

      

      

    

    

      

      

      

      

      

    

    

      

      

      

      

      

    

    

      

      

      

      

      

    

    

      

      

      

      

      

    

  



## 2024-11-17-rev1-000 Test Notebook 

A Jupyter Notebook file 2024-11-17-rev1-000-test.ipynb is available in the repository to test the 2024-11-17-rev1-000 model.