Single Image Super-Resolution Quality Assessment: A Real-World Dataset, Subjective Studies, and An Objective Metric
This is the brief description of RealSRQ dataset and KLTSRQA software. You can change our program as you like and use it for academic, but please refer to its original source and cite our paper.
- Title: Single Image Super-Resolution Quality Assessment: A Real-World Dataset, Subjective Studies, and An Objective Metric
- Publish: IEEE Transactions on Image Processing, 2022
- Authors: Qiuping Jiang, Zhentao Liu, Ke Gu, Feng Shao, Xinfeng Zhang, Hantao Liu, Weisi Lin
- Institution: The School of Information Science and Engineering, Ningbo University
- Paper: 2022-TIP-RealSRQA
Numerous single image super-resolution (SISR) algorithms have been proposed during the past years to reconstruct a high-resolution (HR) image from its low-resolution (LR) observation. However, how to fairly compare the performance of different SISR algorithms/results remains a challenging problem. So far, the lack of comprehensive human subjective study on large-scale real-world SISR datasets and accurate objective SISR quality assessment metrics makes it unreliable to truly understand the performance of different SISR algorithms. We in this paper make efforts to tackle these two issues. Firstly, we construct a real-world SISR quality dataset (i.e., RealSRQ) and conduct human subjective studies to compare the performance of the representative SISR algorithms. Secondly, we propose a new objective metric, i.e., KLTSRQA, based on the Karhunen-Loeve Transform (KLT) to evaluate the quality of SISR images in a no-reference (NR) manner. Experiments on our constructed RealSRQ and the latest synthetic SISR quality dataset (i.e., QADS) have demonstrated the superiority of our proposed KLTSRQA metric, achieving higher consistency with human subjective scores than relevant existing NR image quality assessment (NR-IQA) metrics.
You can download our constructed dataset and proposed software from
BaiduYun Disk: RealSRQ-KLTSRQA-released (key:1121)
Google Drive: RealSRQ-KLTSRQA-released
Important note: Before you use our dataset, please read the README.txt file carefully especially for the data structure of BT-score.mat. I believe this will help you understand our work.
Matlab R2019a
Important Note: other versions may lead to some errors.
If you have any questions of this repo or our paper, please feel free to contact with the authors: jiangqiuping@nbu.edu.cn, zhentaoliu0319@163.com.
If you find this work is useful for you, please cite the following paper:
@ARTICLE{RealSRQ-KLTSRQA,
author={Jiang, Qiuping and Liu, Zhentao and Gu, Ke and Shao, Feng and Zhang, Xinfeng and Liu, Hantao and Lin, Weisi},
journal={IEEE Transactions on Image Processing},
title={Single Image Super-Resolution Quality Assessment: A Real-World Dataset, Subjective Studies, and an Objective Metric},
year={2022},
volume={31},
number={},
pages={2279-2294},
doi={10.1109/TIP.2022.3154588}}