SUPER RESOLUTION: Going from high to very high resolutions

This is the code repository related to Ruben Moyas's MSc in Data Science Master's Thesis.

This repository contains Python code for automating the downloading and setup of two state-of-the-art super-resolution models. It streamlines the data preparation, model configuration, training, validation, and inference phases for high-quality image super-resolution tasks.

This code and its directory system will give an structure and a set of configuration files and scripts ready to be used in order to train, validate and test the following models:

• LIIF
• SR3

Table of Contents

1. Introduction
2. Getting Started
   1. Prerequisites
   2. Installation
3. Usage
4. Structure of the Repository
5. Contributing
6. License
7. Acknowledgements
8. Credits

Introduction

Super-resolution is a class of techniques used in image processing that enhances the resolution of an image. The LIIF and SR3 models are cutting-edge models that use deep learning to achieve remarkable results in super-resolution tasks. This project aims to provide an automated pipeline to facilitate the use and experimentation with these models.

Getting Started

Prerequisites

• Python 3.7 or higher
• PyTorch 1.6 or higher
• CUDA

Installation

1. Clone the repository:

   git clone https://github.com/rmoyav/tfm_super_resolution.git
   

2. Navigate to the repository directory:

   cd tfm_super_resolution
   

3. Install the required dependencies:

   pip install -r requirements.txt
   

Usage

1. Download the dataset and unzip it in a 'data' folder under repository directory: To download the dataset that will be used during the training, validation, and testing of the models, please go to the following URL UC Merced Land Use Dataset and follow the indicated steps. Once you have downloaded the compressed folder with the data, please create a 'data' folder in the root of this repository and extract the contents of the downloaded file inside it.
2. Run the main.py script:

   python ./tools/main.py
   

3. Follow the on-screen instructions to choose from the various options, such as downloading the models, preparing the data, configuring the models, and running training, validation, or inference.

Structure of the Repository (before and after the main script has been used)

• data/: Contains the data used for training and validation.
• liif_script/: Contains the custom infer script for the liif model
• model_config/: Contains the configuration files used for training and validation.
• models/: Contains the liif and SR3 model repositories.
• tools/: Contains the code developed for this project.
• tools/main.py: Main script that automates the entire pipeline through a user-friendly menu.

Contributing

Contributions to this repository will be welcome once the Master's Thesis has been published and defended. Please open an issue to discuss the change or improvement before making a pull request.

License

This project is licensed under the Creative Commons CC0 1.0 Universal license. See the LICENSE file for details.

Acknowledgements

I would like to express my gratitude to my advisors and the Universitat Oberta de Catalunya for their support throughout the development of this project. I also want to acknowledge the original authors and contributors of the LIIF and SR3 models.

Credits

Author: Rubén Moya rmoyav@uoc.edu