Research map / Optimization / Optimizers
Deep Generalized Unfolding Networks for Image Restoration
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 the PGD-unfolded DGUNet instead of MPRNet | Image deraining · Deraining average | PSNR: 32.73 -> 33.06 | Helped |
| Use the PGD-unfolded DGUNet instead of MPRNet | Image deblurring · GoPro | PSNR: 32.66 -> 32.71 | Helped |
| Use DGUNet+ instead of MPRNet | Image deblurring · HIDE | PSNR: 30.96 -> 31.40 | Helped |
| Use DGUNet+ instead of MPRNet | Image denoising · SIDD | PSNR: 39.71 -> 39.91 | Helped |
| Use DGUNet+ instead of MPRNet | Image denoising · DND | PSNR: 39.80 -> 40.12 | Helped |
| Use DGUNet instead of OPINENet+ | Compressive sensing · Set11, 1% CS ratio | PSNR: 20.15 -> 22.09 | Helped |
| Add a ninth stage (ablation) | DGUNet+ · Rain100H | PSNR: 31.06 -> 31.27 | Helped |
| Reduce stages from 7 to 5 (ablation) | DGUNet+ · Rain100H | PSNR: 31.06 -> 30.43 | Hurt |
| Reduce stages from 7 to 3 (ablation) | DGUNet+ · Rain100H | PSNR: 31.06 -> 29.67 | Hurt |
| Use 2 rather than 3 feature-fusion scales (ablation) | DGUNet+ · Rain100H | PSNR: 31.06 -> 30.87 | Hurt |
| Use 1 rather than 3 feature-fusion scales (ablation) | DGUNet+ · Rain100H | PSNR: 31.06 -> 30.68 | Hurt |
| Add the flexible gradient descent module (ablation) | DGUNet+ · Rain100H | PSNR: 30.51 -> 31.06 | Helped |
Closest papers by meaning
- Vision-Language Gradient Descent-driven All-in-One Deep Unfolding Networks (CVPR 2025 · similarity 0.90)
- Deep Unfolding Network for Image Super-Resolution (CVPR 2020 · similarity 0.88)
- Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip Connections (NeurIPS 2016 · similarity 0.88)
- A Restoration Network as an Implicit Prior (ICLR 2024 · similarity 0.88)
- Using Powerful Prior Knowledge of Diffusion Model in Deep Unfolding Networks for Image Compressive Sensing (CVPR 2025 · similarity 0.88)
- Omni-Kernel Network for Image Restoration (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.