Cascaded Degradation-Aware Blind Super-Resolution
Image super-resolution (SR) usually synthesizes degraded low-resolution images with a predefined degradation model for training. Existing SR methods inevitably perform poorly when the true degradation does not follow the predefined degradation, especially in the case of the real world. To tackle thi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/11/5338 |
| _version_ | 1797596703365267456 |
|---|---|
| author | Ding Zhang Ni Tang Dongxiao Zhang Yanyun Qu |
| author_facet | Ding Zhang Ni Tang Dongxiao Zhang Yanyun Qu |
| author_sort | Ding Zhang |
| collection | DOAJ |
| description | Image super-resolution (SR) usually synthesizes degraded low-resolution images with a predefined degradation model for training. Existing SR methods inevitably perform poorly when the true degradation does not follow the predefined degradation, especially in the case of the real world. To tackle this robustness issue, we propose a cascaded degradation-aware blind super-resolution network (CDASRN), which not only eliminates the influence of noise on blur kernel estimation but also can estimate the spatially varying blur kernel. With the addition of contrastive learning, our CDASRN can further distinguish the differences between local blur kernels, greatly improving its practicality. Experiments in various settings show that CDASRN outperforms state-of-the-art methods on both heavily degraded synthetic datasets and real-world datasets. |
| first_indexed | 2024-03-11T02:56:48Z |
| format | Article |
| id | doaj.art-86b172c832254a13a76fffeea68dea33 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T02:56:48Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-86b172c832254a13a76fffeea68dea332023-11-18T08:35:49ZengMDPI AGSensors1424-82202023-06-012311533810.3390/s23115338Cascaded Degradation-Aware Blind Super-ResolutionDing Zhang0Ni Tang1Dongxiao Zhang2Yanyun Qu3School of Information, Xiamen University, Xiamen 361005, ChinaSchool of Science, Jimei University, Xiamen 361021, ChinaSchool of Science, Jimei University, Xiamen 361021, ChinaSchool of Information, Xiamen University, Xiamen 361005, ChinaImage super-resolution (SR) usually synthesizes degraded low-resolution images with a predefined degradation model for training. Existing SR methods inevitably perform poorly when the true degradation does not follow the predefined degradation, especially in the case of the real world. To tackle this robustness issue, we propose a cascaded degradation-aware blind super-resolution network (CDASRN), which not only eliminates the influence of noise on blur kernel estimation but also can estimate the spatially varying blur kernel. With the addition of contrastive learning, our CDASRN can further distinguish the differences between local blur kernels, greatly improving its practicality. Experiments in various settings show that CDASRN outperforms state-of-the-art methods on both heavily degraded synthetic datasets and real-world datasets.https://www.mdpi.com/1424-8220/23/11/5338image super-resolutionmultiple degradation factorsblur kernel estimationcontrast learning |
| spellingShingle | Ding Zhang Ni Tang Dongxiao Zhang Yanyun Qu Cascaded Degradation-Aware Blind Super-Resolution Sensors image super-resolution multiple degradation factors blur kernel estimation contrast learning |
| title | Cascaded Degradation-Aware Blind Super-Resolution |
| title_full | Cascaded Degradation-Aware Blind Super-Resolution |
| title_fullStr | Cascaded Degradation-Aware Blind Super-Resolution |
| title_full_unstemmed | Cascaded Degradation-Aware Blind Super-Resolution |
| title_short | Cascaded Degradation-Aware Blind Super-Resolution |
| title_sort | cascaded degradation aware blind super resolution |
| topic | image super-resolution multiple degradation factors blur kernel estimation contrast learning |
| url | https://www.mdpi.com/1424-8220/23/11/5338 |
| work_keys_str_mv | AT dingzhang cascadeddegradationawareblindsuperresolution AT nitang cascadeddegradationawareblindsuperresolution AT dongxiaozhang cascadeddegradationawareblindsuperresolution AT yanyunqu cascadeddegradationawareblindsuperresolution |