Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging
Currently, deep-learning-based methods have been the most popular super-resolution techniques owing to the improvement of super-resolution performance. However, they are still lack perceptual fine details and thus result in unsatisfying visual quality. This article proposes a novel method for high-q...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9220103/ |
| _version_ | 1818933662382555136 |
|---|---|
| author | Sanyou Zhang Deqiang Cheng Daihong Jiang Qiqi Kou |
| author_facet | Sanyou Zhang Deqiang Cheng Daihong Jiang Qiqi Kou |
| author_sort | Sanyou Zhang |
| collection | DOAJ |
| description | Currently, deep-learning-based methods have been the most popular super-resolution techniques owing to the improvement of super-resolution performance. However, they are still lack perceptual fine details and thus result in unsatisfying visual quality. This article proposes a novel method for high-quality perceptual super-resolution imaging, named SRLRGAN-SN. It aims to recovery visually plausible images with perceptual texture details by using the least squares relativistic generative adversarial network (GAN). The method applies the spectral normalization on the network with the target of enhancing the performance of GAN for super-resolution task. The least squares relativistic discriminator is designed to drive reconstruction images approximating high-quality perceptual manifold. Besides, a novel perceptual loss assembly is proposed to preserve structural texture details as much as possible. Results of experiment show that our method can not only recovery more visually realistic details, but also outperforms other popular methods regarding to quantitative metrics and perceptual evaluations. |
| first_indexed | 2024-12-20T04:51:57Z |
| format | Article |
| id | doaj.art-58acfd3ff7054cfa8c78aad0433d8692 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-20T04:51:57Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-58acfd3ff7054cfa8c78aad0433d86922022-12-21T19:52:50ZengIEEEIEEE Access2169-35362020-01-01818519818520810.1109/ACCESS.2020.30300449220103Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution ImagingSanyou Zhang0https://orcid.org/0000-0001-8372-292XDeqiang Cheng1Daihong Jiang2https://orcid.org/0000-0002-1163-8144Qiqi Kou3https://orcid.org/0000-0003-2873-2636School of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaInformation and Electrical Engineering College, Xuzhou University of Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaCurrently, deep-learning-based methods have been the most popular super-resolution techniques owing to the improvement of super-resolution performance. However, they are still lack perceptual fine details and thus result in unsatisfying visual quality. This article proposes a novel method for high-quality perceptual super-resolution imaging, named SRLRGAN-SN. It aims to recovery visually plausible images with perceptual texture details by using the least squares relativistic generative adversarial network (GAN). The method applies the spectral normalization on the network with the target of enhancing the performance of GAN for super-resolution task. The least squares relativistic discriminator is designed to drive reconstruction images approximating high-quality perceptual manifold. Besides, a novel perceptual loss assembly is proposed to preserve structural texture details as much as possible. Results of experiment show that our method can not only recovery more visually realistic details, but also outperforms other popular methods regarding to quantitative metrics and perceptual evaluations.https://ieeexplore.ieee.org/document/9220103/Generative adversarial networksuper-resolution imagingrelativistic discriminatorperceptual qualityspectral normalization |
| spellingShingle | Sanyou Zhang Deqiang Cheng Daihong Jiang Qiqi Kou Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging IEEE Access Generative adversarial network super-resolution imaging relativistic discriminator perceptual quality spectral normalization |
| title | Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging |
| title_full | Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging |
| title_fullStr | Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging |
| title_full_unstemmed | Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging |
| title_short | Least Squares Relativistic Generative Adversarial Network for Perceptual Super-Resolution Imaging |
| title_sort | least squares relativistic generative adversarial network for perceptual super resolution imaging |
| topic | Generative adversarial network super-resolution imaging relativistic discriminator perceptual quality spectral normalization |
| url | https://ieeexplore.ieee.org/document/9220103/ |
| work_keys_str_mv | AT sanyouzhang leastsquaresrelativisticgenerativeadversarialnetworkforperceptualsuperresolutionimaging AT deqiangcheng leastsquaresrelativisticgenerativeadversarialnetworkforperceptualsuperresolutionimaging AT daihongjiang leastsquaresrelativisticgenerativeadversarialnetworkforperceptualsuperresolutionimaging AT qiqikou leastsquaresrelativisticgenerativeadversarialnetworkforperceptualsuperresolutionimaging |