Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network.
In recent years, deep learning (DL) networks have been widely used in super-resolution (SR) and exhibit improved performance. In this paper, an image quality assessment (IQA)-guided single image super-resolution (SISR) method is proposed in DL architecture, in order to achieve a nice tradeoff betwee...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2020-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0241313&type=printable |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850187772877340672 |
|---|---|
| author | Zhengqiang Xiong Manhui Lin Zhen Lin Tao Sun Guangyi Yang Zhengxing Wang |
| author_facet | Zhengqiang Xiong Manhui Lin Zhen Lin Tao Sun Guangyi Yang Zhengxing Wang |
| author_sort | Zhengqiang Xiong |
| collection | DOAJ |
| description | In recent years, deep learning (DL) networks have been widely used in super-resolution (SR) and exhibit improved performance. In this paper, an image quality assessment (IQA)-guided single image super-resolution (SISR) method is proposed in DL architecture, in order to achieve a nice tradeoff between perceptual quality and distortion measure of the SR result. Unlike existing DL-based SR algorithms, an IQA net is introduced to extract perception features from SR results, calculate corresponding loss fused with original absolute pixel loss, and guide the adjustment of SR net parameters. To solve the problem of heterogeneous datasets used by IQA and SR networks, an interactive training model is established via cascaded network. We also propose a pairwise ranking hinge loss method to overcome the shortcomings of insufficient samples during training process. The performance comparison between our proposed method with recent SISR methods shows that the former achieves a better tradeoff between perceptual quality and distortion measure than the latter. Extensive benchmark experiments and analyses also prove that our method provides a promising and opening architecture for SISR, which is not confined to a specific network model. |
| format | Article |
| id | doaj-art-25f87d1be8704a01ace402650f37b51b |
| institution | OA Journals |
| issn | 1932-6203 |
| language | English |
| publishDate | 2020-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj-art-25f87d1be8704a01ace402650f37b51b2025-08-20T02:16:02ZengPublic Library of Science (PLoS)PLoS ONE1932-62032020-01-011510e024131310.1371/journal.pone.0241313Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network.Zhengqiang XiongManhui LinZhen LinTao SunGuangyi YangZhengxing WangIn recent years, deep learning (DL) networks have been widely used in super-resolution (SR) and exhibit improved performance. In this paper, an image quality assessment (IQA)-guided single image super-resolution (SISR) method is proposed in DL architecture, in order to achieve a nice tradeoff between perceptual quality and distortion measure of the SR result. Unlike existing DL-based SR algorithms, an IQA net is introduced to extract perception features from SR results, calculate corresponding loss fused with original absolute pixel loss, and guide the adjustment of SR net parameters. To solve the problem of heterogeneous datasets used by IQA and SR networks, an interactive training model is established via cascaded network. We also propose a pairwise ranking hinge loss method to overcome the shortcomings of insufficient samples during training process. The performance comparison between our proposed method with recent SISR methods shows that the former achieves a better tradeoff between perceptual quality and distortion measure than the latter. Extensive benchmark experiments and analyses also prove that our method provides a promising and opening architecture for SISR, which is not confined to a specific network model.https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0241313&type=printable |
| spellingShingle | Zhengqiang Xiong Manhui Lin Zhen Lin Tao Sun Guangyi Yang Zhengxing Wang Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. PLoS ONE |
| title | Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. |
| title_full | Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. |
| title_fullStr | Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. |
| title_full_unstemmed | Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. |
| title_short | Single image super-resolution via Image Quality Assessment-Guided Deep Learning Network. |
| title_sort | single image super resolution via image quality assessment guided deep learning network |
| url | https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0241313&type=printable |
| work_keys_str_mv | AT zhengqiangxiong singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork AT manhuilin singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork AT zhenlin singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork AT taosun singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork AT guangyiyang singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork AT zhengxingwang singleimagesuperresolutionviaimagequalityassessmentguideddeeplearningnetwork |