Reproduce “Fidelity-Constrained Anchoring for Black-Box Denoisers” 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
Masaki Satoh
Morpho, Inc. developed a fidelity-constrained anchoring framework that post-processes black-box denoiser outputs by linearly blending them with the noisy input. The approach effectively controls output fidelity to the input image, balancing denoising performance and naturalness, with SSIM-based anchoring demonstrating robustness across varying noise levels.
Key insights
- The local constant-blending assumption effectively enables spatially adaptive and explicit fidelity control over black-box denoisers.
- The inverse SSIM function exhibits positive, monotonically increasing, and convex properties over the valid blending range, allowing for efficient iterative determination of the optimal blending factor.
- SSIM-based anchoring demonstrates higher robustness and consistency across different input noise levels compared to PSNR-based anchoring, which is sensitive to noise.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.