Research map / Image generation / Image restoration
End-to-End Unpaired Image Denoising with Conditional Adversarial Networks
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 |
|---|---|---|---|
| add sharpening (local-mean subtraction) to discriminator input (ablation) | UIDNet · BSD68 sigma=25 | PSNR: 28.05 -> 28.98 | Helped |
| add sharpening to discriminator input (ablation) | UIDNet · SIDD | PSNR: 31.34 -> 32.48 | Helped |
| add sharpening to discriminator input (ablation) | UIDNet · SIDD | SSIM: 0.856 -> 0.897 | Helped |
| add sharpening to discriminator input (ablation) | UIDNet · LDCT NoisyA | PSNR: 26.475 -> 26.694 | Helped |
| add sharpening to discriminator input (ablation) | UIDNet · LDCT NoisyB | PSNR: 21.198 -> 21.247 | Helped |
| binomial instead of default random sample z (ablation) | UIDNet · LDCT noise | learned noise distribution | No clear effect |
| train on unpaired CleanA+NoisyB instead of paired RED-CNN on CleanA/NoisyA | UIDNet vs RED-CNN · LDCT NoisyB test | PSNR: 21.723 -> 22.315 | Helped |
| add NoisyA to NoisyB training noisy data (ablation) | UIDNet · LDCT NoisyB test | PSNR: 22.315 -> 22.578 | Helped |
Closest papers by meaning
- GAN2GAN: Generative Noise Learning for Blind Denoising with Single Noisy Images (ICLR 2021 · similarity 0.92)
- Universal Denoising Networks : A Novel CNN Architecture for Image Denoising (CVPR 2018 · similarity 0.91)
- Image Blind Denoising With Generative Adversarial Network Based Noise Modeling (CVPR 2018 · similarity 0.91)
- Learning to Generate Realistic Noisy Images via Pixel-level Noise-aware Adversarial Training (NeurIPS 2021 · similarity 0.90)
- Toward Convolutional Blind Denoising of Real Photographs (CVPR 2019 · similarity 0.90)
- Robust And Interpretable Blind Image Denoising Via Bias-Free Convolutional Neural Networks (ICLR 2020 · similarity 0.90)
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.