Research map / Image generation / Image restoration
Masked Image Training for Generalizable Deep Image Denoising
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 |
|---|---|---|---|
| Apply masked training instead of standard training | SwinIR · Kodak24 mixture noise level 1 | PSNR: 29.23 -> 30.31 | Helped |
| Apply masked training instead of standard training | SwinIR · Synthetic ISP noise | LPIPS: 0.3083 -> 0.2368 | Helped |
| Apply masked training instead of standard training | SwinIR · Monte Carlo rendered images, 128 samples per pixel | PSNR: 29.96 -> 30.62 | Helped |
| Apply masked training instead of standard training | SwinIR · Monte Carlo rendered images, 64 samples per pixel | PSNR: 26.28 -> 28.25 | Helped |
| Pre-train on clean images with masked training before SIDD fine-tuning | SwinIR · SIDD, 10-image test set | PSNR: 38.36 -> 39.11 | Helped |
| Use a random input mask ratio of 75–85% instead of a fixed ratio (ablation) | Masked training · Mixture noise on CBSD68 | PSNR: 29.17 -> 30.04 | Helped |
| Increase the fixed input mask ratio from 85% to 95% (ablation) | Masked training · Mixture noise on CBSD68 | PSNR: 29.44 -> 19.60 | Hurt |
| Add the input mask (ablation) | Masked training · Unseen noise types | denoising generalization | Helped |
| Add the attention mask (ablation) | Masked training · Unmasked test inputs | quantitative denoising performance | Helped |
Closest papers by meaning
- Robust And Interpretable Blind Image Denoising Via Bias-Free Convolutional Neural Networks (ICLR 2020 · similarity 0.91)
- Random Is All You Need: Random Noise Injection on Feature Statistics for Generalizable Deep Image Denoising (ICLR 2025 · similarity 0.91)
- High-Quality Self-Supervised Deep Image Denoising (NeurIPS 2019 · similarity 0.89)
- Masked Pre-training Enables Universal Zero-shot Denoiser (NeurIPS 2024 · similarity 0.88)
- Zero-Shot Noise2Mean: Gap Minimization for Efficient Denoising from a Single Noisy Image (AAAI 2025 · similarity 0.86)
- Classic Video Denoising in a Machine Learning World: Robust, Fast, and Controllable (CVPR 2025 · similarity 0.86)
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.