Reproduce “Rethinking Noise Synthesis and Modeling in Raw Denoising” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yi Zhang, Hongwei Qin, Xiaogang Wang, Hongsheng Li
The authors present a raw image denoising method that samples signal-independent noise from target-camera dark frames and models shot noise with a Poisson distribution. Pattern-aligned patch sampling preserves spatial and Bayer structure, while high-bit reconstruction addresses dark-frame quantization. On SonyA7S2 at exposure ratio 100, PAP + HB reaches 47.90 dB PSNR and 0.973 SSIM, but results remain camera- and setting-specific.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.