Research map / Image generation / Image restoration
Divergence-Free Neural Networks with Application to Image Denoising
Sébastien Herbreteau, Etienne Meunier
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 |
|---|---|---|---|
| use CENSURE instead of UNSURE when sigma varies | CENSURE · Kodak24 color, sigma=25 | PSNR (dB): 22.03 -> 32.05 | Improved |
| use CENSURE instead of Noise2Self | CENSURE · Kodak24 color, sigma=25 | PSNR (dB): 31.90 -> 32.05 | Improved |
| use CENSURE instead of Noise2Self | CENSURE · Set12 gray, constant unknown sigma=25 | PSNR (dB): 29.55 -> 29.81 | Improved |
| use Neighbor2Neighbor instead of CENSURE | Neighbor2Neighbor · CBSD68 color, sigma=25 | PSNR (dB): 30.83 -> 31.29 | Improved |
| use UNSURE instead of CENSURE with constant unknown sigma | UNSURE · Set12 gray, constant unknown sigma=25 | PSNR (dB): 29.81 -> 29.84 | Improved |
| increase number of terms K' from 2 to 8 (ablation) | CENSURE · Set12 gray, constant unknown sigma=25 | PSNR (dB): 29.44 -> 29.81 | Improved |
| increase K' from 2 to 8 (ablation) | CENSURE · Inference time | seconds: 0.024 -> 0.072 | Hurt |
| use CENSURE instead of supervised network | CENSURE · Inference time | seconds: 0.007 -> 0.072 | Hurt |
| use tau=1e-4 instead of 1e-2 in UNSURE (ablation) | UNSURE · Kodak24 color, sigma=25 | PSNR (dB): 22.03 -> 22.01 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.
- Neighbor2Neighbor — Neighbor2Neighbor: Self-Supervised Denoising From Single Noisy Images (CVPR 2021) · better on 0 of 4 shared results
- Noise2Score — Noise2Score: Tweedie’s Approach to Self-Supervised Image Denoising without Clean Images (NeurIPS 2021) · better on 1 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Convexity-Aware Noise Calibration: A Self-Supervised Framework for Noise-Level-Unknown Image Denoising (CVPR 2026 · similarity 0.89)
- Robust And Interpretable Blind Image Denoising Via Bias-Free Convolutional Neural Networks (ICLR 2020 · similarity 0.89)
- Unsupervised Image Restoration Using Partially Linear Denoisers (TPAMI 2021 · similarity 0.89)
- Universal Denoising Networks : A Novel CNN Architecture for Image Denoising (CVPR 2018 · similarity 0.89)
- Robust Diffusion Models via Divergence-Induced Weighted Denoising (NeurIPS 2026 · similarity 0.88)
- Generalization in diffusion models arises from geometry-adaptive harmonic representations (ICLR 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.
