I Q A Py Torch Weights

Introduction

This repository contains PyTorch model weights for Image Quality Assessment (IQA) algorithms. These weights support a variety of IQA models and metrics, facilitating both full-reference and no-reference image quality evaluation tasks.

Architecture

The repository includes model weights sourced from two origins:

1. Models trained and validated by the contributing team.
2. Official weights sourced from the original model repositories.

Training

The weights were created through a combination of internal training and validation efforts as well as collecting official weights from existing repositories. The performance of these models aims to reflect the original research but may not always match the results stated in the papers.

Guide: Running Locally

1. Clone the IQA-PyTorch repository.
2. Follow the usage instructions provided in the documentation of the IQA-PyTorch repository for integrating the weights with specific models.
3. For optimal performance, consider using cloud GPU providers such as AWS, Google Cloud, or Azure.

License

The model weights are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (cc-by-nc-sa-4.0). Please adhere to this license and cite the original papers when utilizing these weights in your work.