---
license: mit
---

# Overview

<p align="center">
  <img src="https://avatars.githubusercontent.com/u/12619994?s=200&v=4" width="150">
</p>

<!-- -------------------------------------------------------------------------------- -->

AIPQ is a full-reference image quality assessment method. Three model checkpoints are provided in this repository, these models are to be used together with [this github repo](https://github.com/huawei-noah/noah-research/tree/master/aipq)

## Citation

Please cite the following [paper](https://bmvc2022.mpi-inf.mpg.de/0244.pdf) when using our code or model:

``` bibtex
@inproceedings{thong2022content,
  title={Content-Diverse Comparisons improve {IQA}},
  author={Thong, William and Costa Pereira, Jose and Parisot, Sarah and Leonardis, Ales and McDonagh, Steven},
  booktitle={British Machine Vision Conference},
  year={2022}
}
```