English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

AdaptivePnP: Fast Auto-Tuned Plug-and-Play for Image Deconvolution

Forum topic · 小凯 · 2026-11-09

Summary

Plug-and-play (PnP) image reconstruction methods alternate a data-fitting step with a denoiser, but their performance depends on a step size (gamma) that is usually grid-searched on validation images and a noise level (sigma) that is assumed known. AdaptivePnP, by Duba-Sullivan, Reid, Bouman, and Buzzard (arXiv:2611.09845), removes both requirements by jointly estimating sigma and gamma at every iteration directly from the data-fitting residual. The authors apply the PnP virtual model to the data-fitting proximal map and prove the residual is zero-mean Gaussian with covariance parameterized by gamma and sigma. Because the residual carries little information about one parameter when the other dominates, they introduce an identifiable maximum likelihood (IML) estimator that constrains the ML estimate to a window where the normalized Cramer-Rao bound of each parameter stays within a user-specified tolerance. For the virtual model this window is an interval of the ratio rho = gamma/sigma, so IML reduces to a 1D search. Applied to two standard PnP methods, AdaptivePnP matches or outperforms baselines tuned with ground-truth sigma on deblurring and super-resolution.

Overview

Field: Computer Vision Authors: Haley Duba-Sullivan, Emma J. Reid, Charles A. Bouman, Gregery T. Buzzard Published: 2026-11-06 arXiv: 2611.09845

Summary

Plug-and-play (PnP) methods reconstruct images by alternating a data-fitting step with an image denoiser. Their performance depends on two parameters: the step size \(\gamma\) and the measurement noise level \(\sigma\). In practice, \(\gamma\) is often tuned by grid search over validation images and must be retuned whenever the problem setting changes, and \(\sigma\) is assumed known or estimated separately.

AdaptivePnP is a PnP framework that jointly estimates \(\sigma\) and \(\gamma\) at every iteration from the data-fitting residual. To do so, the authors apply the PnP virtual model to the data-fitting proximal map and show that the residual is zero-mean Gaussian with covariance parameterized by \(\gamma\) and \(\sigma\).

However, when one parameter is much larger than the other, the residual carries little information about the smaller one, and using a plain maximum likelihood (ML) estimator can degrade PnP iterations. The authors therefore introduce an identifiable maximum likelihood (IML) estimator, which restricts the ML estimate to a window where the normalized Cramér-Rao bound of each parameter is bounded by a user-specified tolerance. For the virtual model, this window is an interval of the ratio \(\rho = \gamma/\sigma\), so computing the IML estimate reduces to a one-dimensional search over \(\rho\).

Results

  • AdaptivePnP is applied to two standard PnP methods, producing algorithms that require neither tuning of \(\gamma\) nor knowledge of \(\sigma\).
  • On image deblurring and super-resolution tasks, AdaptivePnP matches or exceeds PnP baselines tuned with ground-truth \(\sigma\).

Original Abstract (excerpt)

> Plug-and-play (PnP) methods reconstruct images by alternating a data-fitting step with an image denoiser. Their performance depends on two parameters: the step size \(\gamma\) and the measurement noise level \(\sigma\). In practice, \(\gamma\) is often tuned by grid search over validation images and must be retuned whenever the problem setting changes, and \(\sigma\) is assumed known or estimated separately. We propose AdaptivePnP, a PnP framework that jointly estimates \(\sigma\) and \(\gamma\) at every iteration from the data-fitting residual. To do so, we apply the PnP virtual model to the data-fitting proximal map and show that the residual is zero-mean Gaussian with covariance parameterized by \(\gamma\) and \(\sigma\). However, when one parameter is much larger than the other, the residual carries little information about the smaller one ...

---

*Collected automatically on 2026-11-10. Source: arXiv:2611.09845*

Tags

#plug-and-play#image-deconvolution#hyperparameter-tuning#maximum-likelihood#deblurring#super-resolution#computational-imaging

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178636991