Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric
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. S...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/162755 |
| _version_ | 1811679095569776640 |
|---|---|
| author | Jiang, Qiuping Liu, Zhentao Gu, Ke Shao, Feng Zhang, Xinfeng Liu, Hantao Lin, Weisi |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Jiang, Qiuping Liu, Zhentao Gu, Ke Shao, Feng Zhang, Xinfeng Liu, Hantao Lin, Weisi |
| author_sort | Jiang, Qiuping |
| collection | NTU |
| description | 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-Loéve 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. The dataset and the code will be made available at https://github.com/Zhentao-Liu/RealSRQ-KLTSRQA. |
| first_indexed | 2024-10-01T03:03:42Z |
| format | Journal Article |
| id | ntu-10356/162755 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:03:42Z |
| publishDate | 2022 |
| record_format | dspace |
| spelling | ntu-10356/1627552022-11-08T04:10:06Z Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric Jiang, Qiuping Liu, Zhentao Gu, Ke Shao, Feng Zhang, Xinfeng Liu, Hantao Lin, Weisi School of Computer Science and Engineering Engineering::Computer science and engineering Single Image Super-Resolution Image Quality Assessment 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-Loéve 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. The dataset and the code will be made available at https://github.com/Zhentao-Liu/RealSRQ-KLTSRQA. This work was supported in part by the Zhejiang Natural Science Foundation under Grant LR22F020002; in part by the Natural Science Foundation of China under Grant 61901236, Grant 62071261, Grant 62076013, Grant 62071449, and Grant U20A20184; in part by the Beijing Natural Science Foundation under Grant JQ21014; in part by the Fundamental Research Funds for the Provincial Universities of Zhejiang under Grant SJLZ2020003; and in part by the Fundamental Research Funds for the Central Universities. 2022-11-08T04:10:05Z 2022-11-08T04:10:05Z 2022 Journal Article Jiang, Q., Liu, Z., Gu, K., Shao, F., Zhang, X., Liu, H. & Lin, W. (2022). Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric. IEEE Transactions On Image Processing, 31, 2279-2294. https://dx.doi.org/10.1109/TIP.2022.3154588 1057-7149 https://hdl.handle.net/10356/162755 10.1109/TIP.2022.3154588 35239481 2-s2.0-85125730905 31 2279 2294 en IEEE Transactions on Image Processing © 2022 IEEE. All rights reserved. |
| spellingShingle | Engineering::Computer science and engineering Single Image Super-Resolution Image Quality Assessment Jiang, Qiuping Liu, Zhentao Gu, Ke Shao, Feng Zhang, Xinfeng Liu, Hantao Lin, Weisi Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title | Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title_full | Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title_fullStr | Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title_full_unstemmed | Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title_short | Single image super-resolution quality assessment: a real-world dataset, subjective studies, and an objective metric |
| title_sort | single image super resolution quality assessment a real world dataset subjective studies and an objective metric |
| topic | Engineering::Computer science and engineering Single Image Super-Resolution Image Quality Assessment |
| url | https://hdl.handle.net/10356/162755 |
| work_keys_str_mv | AT jiangqiuping singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT liuzhentao singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT guke singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT shaofeng singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT zhangxinfeng singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT liuhantao singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric AT linweisi singleimagesuperresolutionqualityassessmentarealworlddatasetsubjectivestudiesandanobjectivemetric |