An Online Plug-and-Play Algorithm for Regularized Image Reconstruction
This work addresses the problem of handling large datasets in imaging inverse problems for researchers and practitioners, but it is incremental as it builds on existing PnP methods.
The paper tackles the scalability of plug-and-play priors for image reconstruction by introducing an online algorithm based on ISTA that uses subsets of measurements per iteration, enabling application to large datasets, with theoretical convergence analysis and simulations in diffraction tomography.
Plug-and-play priors (PnP) is a powerful framework for regularizing imaging inverse problems by using advanced denoisers within an iterative algorithm. Recent experimental evidence suggests that PnP algorithms achieve state-of-the-art performance in a range of imaging applications. In this paper, we introduce a new online PnP algorithm based on the iterative shrinkage/thresholding algorithm (ISTA). The proposed algorithm uses only a subset of measurements at every iteration, which makes it scalable to very large datasets. We present a new theoretical convergence analysis, for both batch and online variants of PnP-ISTA, for denoisers that do not necessarily correspond to proximal operators. We also present simulations illustrating the applicability of the algorithm to image reconstruction in diffraction tomography. The results in this paper have the potential to expand the applicability of the PnP framework to very large and redundant datasets.