Adaptivepnp automatically tunes parameters for faster image restoration
AdaptivePnP: Fast Auto-Tuned Plug-and-Play for Deconvolution
Computer Vision and Pattern Recognition
Summary
Restoring clear images from blurry or noisy ones often needs careful fine-tuning of certain settings, which can be slow and tricky. The authors propose AdaptivePnP, a new method that adjusts these settings automatically during the restoration process by looking at how well the current guess matches the observed data. This approach avoids trial-and-error tuning and works well across different tasks like sharpening blurred images or increasing image resolution. Their results show that AdaptivePnP performs as well or better than traditional methods that need manual tuning.
What this means in practice
- •For image processing engineers: Implement image deblurring and super-resolution algorithms that automatically adjust parameters without manual tuning for varied imaging conditions.
- •For medical imaging technicians: Improve quality of medical images like MRIs or CT scans by applying adaptive deconvolution methods that do not require manual parameter setting for each scan.
Authors
Haley Duba-Sullivan, Emma J. Reid, Charles A. Bouman, Gregery T. Buzzard
Abstract
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 $γ$ and the measurement noise level $σ$. In practice, $γ$ is often tuned by grid search over validation images and must be retuned whenever the problem setting changes, and $σ$ is assumed known or estimated separately. We propose AdaptivePnP, a PnP framework that jointly estimates $σ$ and $γ$ 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 $γ$ and $σ$. However, when one parameter is much larger than the other, the residual carries little information about the smaller one and using the maximum likelihood (ML) estimator can degrade the PnP iterations. We therefore introduce the identifiable maximum likelihood (IML) estimator, which restricts the ML estimator to an identifiable window in which the normalized Cramér--Rao bound of every parameter is bounded by a user-specified tolerance. For the virtual model, this window is an interval of ratios $ρ= γ/σ$, so computing the IML estimate reduces to a one-dimensional search over $ρ$. We apply AdaptivePnP to two standard PnP methods, yielding algorithms that require neither a tuned $γ$ nor a known $σ$. On image deblurring and super-resolution, AdaptivePnP matches or outperforms PnP baselines tuned with the true $σ$. Code is available at https://github.com/hdubasullivan/AdaptivePnP.