Research map / Image generation / Super-resolution
Deep Unfolding Network for Image Super-Resolution
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use end-to-end trained USRNet instead of IRCNN | SISR on BSD68, ×2, noise 0, kernel column 1 · BSD68, ×2, noise 0, kernel column 1 | PSNR (dB): 30.55 vs. 29.60 | Helped |
| Use end-to-end trained USRNet instead of IRCNN | SISR on BSD68, ×3, noise 0, kernel column 1 · BSD68, ×3, noise 0, kernel column 1 | PSNR (dB): 27.16 vs. 25.97 | Helped |
| Use end-to-end trained USRNet instead of IRCNN | SISR on BSD68, ×3, noise 2.55, kernel column 1 · BSD68, ×3, noise 2.55, kernel column 1 | PSNR (dB): 26.99 vs. 25.70 | Helped |
| Use end-to-end trained USRNet instead of IRCNN | SISR on BSD68, ×3, noise 7.65, kernel column 1 · BSD68, ×3, noise 7.65, kernel column 1 | PSNR (dB): 26.45 vs. 24.58 | Helped |
| Use end-to-end trained USRNet instead of IRCNN | SISR on BSD68, ×4, noise 0, kernel column 1 · BSD68, ×4, noise 0, kernel column 1 | PSNR (dB): 25.30 vs. 23.99 | Helped |
| Reduce USRNet iterations from 8 to 5 (ablation) | USRNet · Gaussian blur kernels | PSNR (dB): decreases about 0.1dB | Hurt |
| Reduce USRNet iterations from 8 to 5 (ablation) | USRNet · Motion blur kernels | PSNR (dB): decreases about 0.3dB | Hurt |
| Fine-tune USRNet with weighted L1, VGG perceptual, and relativistic adversarial losses | USRNet, scale factor 4 · Visual results on scale factor 4 | visual quality | Helped |
| Use USRNet with approximated bicubic kernels despite no training on those kernels | USRNet · Bicubic degradation | PSNR (dB) | Helped |
Closest papers by meaning
- Learning a Single Convolutional Super-Resolution Network for Multiple Degradations (CVPR 2018 · similarity 0.89)
- Noise-Free Optimization in Early Training Steps for Image Super-resolution (AAAI 2024 · similarity 0.89)
- Closed-Loop Matters: Dual Regression Networks for Single Image Super-Resolution (CVPR 2020 · similarity 0.89)
- Deep Generalized Unfolding Networks for Image Restoration (CVPR 2022 · similarity 0.88)
- Image Formation Model Guided Deep Image Super-Resolution (AAAI 2020 · similarity 0.88)
- ResDiff: Combining CNN and Diffusion Model for Image Super-resolution (AAAI 2024 · similarity 0.87)
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.