Research map / Image generation / Image restoration
Multi-Scale Weighted Nuclear Norm Image Restoration
Read the paperCheck 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 whole-image weighted nuclear norm regularizer with multi-scale patch groups | Our · Set5 deblurring, four blur/noise settings | PSNR (dB): 33.54 vs 33.30; 33.02 vs 32.85; 31.83 vs 31.70; 31.13 vs 30.97 | Helped |
| Use the whole-image weighted nuclear norm regularizer with multi-scale patch groups | Our · Set NCSR deblurring, four blur/noise settings | PSNR (dB): 31.06 vs 30.76; 30.38 vs 30.22; 30.22 vs 30.10; 29.34 vs 29.21 | Helped |
| Use the whole-image weighted nuclear norm regularizer with multi-scale patch groups | Our · BSD100 deblurring, four blur/noise settings | PSNR (dB): 29.18 vs 28.98; 28.75 vs 28.58; 28.43 vs 28.25; 27.74 vs 27.57 | Helped |
| Use the whole-image weighted nuclear norm regularizer with multi-scale patch groups | Our · Set5 inpainting, 25%, 50%, and 75% missing pixels | PSNR (dB): 41.85 vs 41.51; 36.77 vs 36.75; 32.00 vs 31.92 | Helped |
| Use the whole-image weighted nuclear norm regularizer with multi-scale patch groups | Our · Set NCSR inpainting, 25%, 50%, and 75% missing pixels | PSNR (dB): 41.29 vs 40.55; 35.77 vs 35.44; 30.06 vs 30.01 | Helped |
| Add multi-scale patch recurrence (ablation) | Our · Set5 deblurring, Gaussian blur, σn = √2 | PSNR (dB): 33.17 -> 33.54 | Helped |
| Add multi-scale patch recurrence (ablation) | Our · Set5 deblurring, uniform blur, σn = √2 | PSNR (dB): 31.57 -> 31.83 | Helped |
Closest papers by meaning
- AdaIR: Adaptive All-in-One Image Restoration via Frequency Mining and Modulation (ICLR 2025 · similarity 0.85)
- Nuclear Norm Regularization for Deep Learning (NeurIPS 2024 · similarity 0.84)
- Residual Denoising Diffusion Models (CVPR 2024 · similarity 0.83)
- IR-CM: The Fast and General-purpose Image Restoration Method Based on Consistency Model (NeurIPS 2024 · similarity 0.83)
- Resfusion: Denoising Diffusion Probabilistic Models for Image Restoration Based on Prior Residual Noise (NeurIPS 2024 · similarity 0.83)
- SeNM-VAE: Semi-Supervised Noise Modeling with Hierarchical Variational Autoencoder (CVPR 2024 · similarity 0.83)
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.