Research map / Image generation / GANs
Unsupervised Image Denoising in Real-World Scenarios via Self-Collaboration Parallel Generative Adversarial Branches
Xin Lin, Chao Ren, Xiao Liu, Jie Huang, Yinjie Lei
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 the NE module (noise extraction) to the generator input (ablation) | GAN-based unsupervised denoiser · SIDD Benchmark | PSNR: 33.26 -> 33.45 | Improved |
| add BGM loss (ablation) | GAN-based unsupervised denoiser · SIDD Benchmark | PSNR: 33.14 -> 33.26 | Improved |
| add self-synthesis constraint (ablation) | V3 · SIDD Benchmark | PSNR: 33.45 -> 34.27 | Improved |
| use parallel two-branch design instead of single branch (ablation) | SGabNet · SIDD Benchmark | PSNR: 34.27 -> 34.67 | Improved |
| add self-collaboration iterations (ablation) | PGabNet (DeamNet) · SIDD Benchmark | PSNR: 34.67 -> 36.53 | Improved |
| add self-collaboration to DnCNN denoiser (ablation) | PGabNet · SIDD Benchmark | PSNR: 30.56 -> 31.52 | Improved |
| add self-collaboration to UNet denoiser (ablation) | PGabNet · SIDD Benchmark | PSNR: 34.63 -> 36.00 | Improved |
| add self-collaboration to DBSNL denoiser (ablation) | PGabNet · SIDD Benchmark | PSNR: 34.29 -> 35.86 | Improved |
| later self-collaboration iterations (ablation) | SCPGabNet · SIDD Benchmark | PSNR gain per iteration: first iteration >0.5dB; last 0.02dB | No clear effect |
| apply self-collaboration to DnCNN (ablation) | PGabNet · DND Benchmark | SSIM: 0.7876 -> 0.7892 | No clear effect |
Closest papers by meaning
- GAN2GAN: Generative Noise Learning for Blind Denoising with Single Noisy Images (ICLR 2021 · similarity 0.92)
- End-to-End Unpaired Image Denoising with Conditional Adversarial Networks (AAAI 2020 · similarity 0.92)
- Dual Adversarial Network: Toward Real-world Noise Removal and Noise Generation (ECCV 2020 · similarity 0.92)
- You Only Sample Once: Taming One-Step Text-to-Image Synthesis by Self-Cooperative Diffusion GANs (ICLR 2025 · similarity 0.88)
- Zero-Shot Noise2Mean: Gap Minimization for Efficient Denoising from a Single Noisy Image (AAAI 2025 · similarity 0.88)
- sRGB Real Noise Modeling via Noise-Aware Sampling with Normalizing Flows (ICLR 2024 · similarity 0.88)
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.
