Research map / Image generation / Super-resolution
KXNet: A Model-Driven Deep Neural Network for Blind Super-Resolution
Jiahong Fu, Hong Wang, Qi Xie, Qian Zhao, Deyu Meng, Zongben Xu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add 5 unfolded stages (S=0 to S=5) (ablation) | KXNet · Set14 | PSNR: 25.74 -> 29.91 | Improved |
| increase stages from 5 to 10 (ablation) | KXNet · Set14 | PSNR: 29.91 -> 30.57 | Improved |
| increase stages from 10 to 17 (ablation) | KXNet · Set14 | PSNR: 30.57 -> 30.96 | Improved |
| increase stages from 17 to 19 (ablation) | KXNet · Set14 | PSNR: 30.96 -> 31.14 | Improved |
| increase stages from 19 to 21 (ablation) | KXNet · Set14 | PSNR: 31.14 -> 31.14 | No clear effect |
| increase stages from 19 to 21 (ablation) | KXNet · Set14 | inference time (s): 0.59 -> 0.64 | Hurt |
| provide ground-truth blur kernel to X-net (ablation) | KXNet · Set14 | PSNR: 31.97 -> 32.85 | Improved |
| provide ground-truth blur kernel to DAN (ablation) | DAN · Set14 | PSNR: 31.97 -> 32.67 | Improved |
| KXNet instead of DAN under complex degradation (noise 0) | blind SR · Urban100 (Setting 2, x3) | PSNR: 25.82 -> 26.37 | Improved |
| KXNet instead of DAN with isotropic blur | blind SR · Urban100 (Setting 1, x4) | PSNR: 25.95 -> 26.18 | Improved |
| KXNet instead of DAN | blind SR · Set5 (Setting 1, x4) | PSNR: 31.96 -> 31.94 | No clear effect |
| KXNet instead of DAN for inference speed | blind SR · Set5 (Setting 2, x2) | inference time (s): 0.52 -> 0.38 | Improved |
Later papers that compared with this method
1 later accepted paper reports KXNet as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Rotation Equivariant Proximal Operator for Deep Unfolding Methods in Image Restoration (TPAMI 2024 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning the Non-Differentiable Optimization for Blind Super-Resolution (CVPR 2021 · similarity 0.91)
- Blind Super-Resolution With Iterative Kernel Correction (CVPR 2019 · similarity 0.90)
- Spectrum-to-Kernel Translation for Accurate Blind Image Super-Resolution (NeurIPS 2021 · similarity 0.89)
- Deep Plug-And-Play Super-Resolution for Arbitrary Blur Kernels (CVPR 2019 · similarity 0.89)
- Uncertainty Learning in Kernel Estimation for Multi-stage Blind Image Super-Resolution (ECCV 2022 · similarity 0.89)
- Navigating Beyond Dropout: An Intriguing Solution towards Generalizable Image Super Resolution (CVPR 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
