Research map / Image generation / Super-resolution
Exploiting Raw Images for Real-Scene Super-Resolution
Xiangyu Xu, Yongrui Ma, Wenxiu Sun, Ming–Hsuan Yang
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 color correction module (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 22.04 -> 31.06 | Improved |
| add raw input branch alongside color image (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 30.21 -> 31.06 | Improved |
| add raw input branch (ablation) | Ours · Synthetic raw SR (non-blind) | PSNR: 30.65 -> 32.07 | Improved |
| feed both raw and color images into RDN | RDN · Synthetic raw SR (blind) | PSNR: 29.93 -> 30.35 | Improved |
| feed both raw and color images into SID | SID · Synthetic raw SR (blind) | PSNR: 21.87 -> 29.55 | Improved |
| train with proposed data generation pipeline instead of previous data (ablation) | RDN · Real images | DIIVINE: 59.11 -> 54.80 | Improved |
| train with proposed data generation pipeline instead of previous data (ablation) | Ours · Real images | DIIVINE: 59.01 -> 46.83 | Improved |
| synthesize degradation in raw space rather than nonlinear color space (ablation) | Ours · Real images | DIIVINE: 47.51 -> 46.83 | Improved |
| add encoder-decoder structure (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 30.68 -> 31.06 | Improved |
| add channel attention (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 30.90 -> 31.06 | Improved |
| add feature fusion to learning-based guided filter (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 30.82 -> 31.06 | Improved |
| replace dense block with DCA block (with single-scale fusion) (ablation) | Ours · Synthetic raw SR (blind) | PSNR: 30.79 -> 30.93 | Improved |
Closest papers by meaning
- Towards Real Scene Super-Resolution With Raw Images (CVPR 2019 · similarity 0.95)
- Image Super-Resolution Using Deep Convolutional Networks (TPAMI 2016 · similarity 0.92)
- Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network (CVPR 2016 · similarity 0.91)
- Accurate Image Super-Resolution Using Very Deep Convolutional Networks (CVPR 2016 · similarity 0.91)
- Learning To Zoom Inside Camera Imaging Pipeline (CVPR 2022 · similarity 0.90)
- Towards Realistic Data Generation for Real-World Super-Resolution (ICLR 2025 · similarity 0.89)
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.
