Research map / Image generation / Image restoration
Denoising the Deep Sky: Physics-Based CCD Noise Formation for Astronomical Imaging
Shuhong Liu, Xining Ge, Ziying Gu, Quanfeng Xu, Ziteng Cui, Lin Gu, Xuangeng Chu, Jun Liu, Dong Li, Tatsuya Harada
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 PRNU modeling to noise synthesis (ablation) | Ours · MuSCAT-3 G real | FLUX-SNR: 35.51 -> 41.21 | Improved |
| add dark-current noise modeling (ablation) | Ours · MuSCAT-3 G real | FLUX-SNR: 38.21 -> 41.21 | Improved |
| add photon shot noise modeling (ablation) | Ours · MuSCAT-3 G real | FLUX-SNR: 32.16 -> 41.21 | Improved |
| add read-noise (fixed pattern) modeling (ablation) | Ours · MuSCAT-3 G real | FLUX-SNR: 31.69 -> 41.21 | Improved |
| add hot-pixel and cosmic-ray impulse noise (ablation) | Ours · MuSCAT-3 G real | FLUX-SNR: 40.88 -> 41.21 | Improved |
| add hot-pixel and cosmic-ray impulse noise (ablation) | Ours · MuSCAT-3 G real | NMAD: 2.46 -> 2.58 | Hurt |
| add PRNU modeling (ablation) | Ours · MuSCAT-4 G zero-shot | FLUX-SNR: 35.17 -> 42.07 | Improved |
| add photon shot noise modeling (ablation) | Ours · MuSCAT-4 G zero-shot | FLUX-SNR: 31.51 -> 42.07 | Improved |
| add photon shot noise modeling (ablation) | Ours · MuSCAT-3 G real | NMAD: 5.72 -> 2.58 | Improved |
| stack 10-20 exposures into Mean high-SNR base instead of raw frames | raw frames · MuSCAT-3 background noise | NMAD: 10.50 -> 3.12 | Improved |
| use physics-based noise synthesis in place of Paired training on stacked data | Paired · MuSCAT-3 avg | FLUX-SNR: 27.38 -> 39.91 | Improved |
| use physics-based synthesis instead of PNNP | PNNP · MuSCAT-4 zero-shot avg | PSNR: 58.13 -> 58.85 | Improved |
Closest papers by meaning
- Physics-based Noise Modeling for Extreme Low-light Photography (TPAMI 2021 · similarity 0.87)
- Listening to the noise: Blind Denoising with Gibbs Diffusion (ICML 2024 · similarity 0.84)
- FluxFlow: Conservative Flow-Matching for Astronomical Image Super-Resolution (NeurIPS 2026 · similarity 0.83)
- Ditch the Denoiser: Emergence of Noise Robustness in Self-Supervised Learning from Data Curriculum (NeurIPS 2025 · similarity 0.83)
- A New Statistical Model of Star Speckles for Learning to Detect and Characterize Exoplanets in Direct Imaging Observations (CVPR 2025 · similarity 0.83)
- Noise Modeling in One Hour: Minimizing Preparation Efforts for Self-supervised Low-Light RAW Image Denoising (CVPR 2025 · 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.
