FISTA Iterates Converge Linearly for Denoiser-Driven Regularization

TL;DR

This paper proves that certain denoiser-driven algorithms for image reconstruction, specifically PnP-FISTA and RED-APG, converge linearly under specific conditions, providing theoretical guarantees for their effectiveness.

Contribution

It establishes the first global linear convergence guarantees for PnP-FISTA and RED-APG algorithms in linear inverse problems with linear denoisers.

Findings

Linear convergence proven for PnP-FISTA and RED-APG

Spectral analysis used to establish convergence guarantees

Applicable to linear inverse problems with linear denoisers

Abstract

The effectiveness of denoising-driven regularization for image reconstruction has been widely recognized. Two prominent algorithms in this area are Plug-and-Play ($\texttt{PnP}$) and Regularization-by-Denoising ($\texttt{RED}$). We consider two specific algorithms $\texttt{PnP-FISTA}$ and $\texttt{RED-APG}$, where regularization is performed by replacing the proximal operator in the $\texttt{FISTA}$ algorithm with a powerful denoiser. The iterate convergence of $\texttt{FISTA}$ is known to be challenging with no universal guarantees. Yet, we show that for linear inverse problems and a class of linear denoisers, global linear convergence of the iterates of $\texttt{PnP-FISTA}$ and $\texttt{RED-APG}$ can be established through simple spectral analysis.