A fast patch-dictionary method for whole image recovery
From MaRDI portal
Publication:326397
Abstract: Various algorithms have been proposed for dictionary learning. Among those for image processing, many use image patches to form dictionaries. This paper focuses on whole-image recovery from corrupted linear measurements. We address the open issue of representing an image by overlapping patches: the overlapping leads to an excessive number of dictionary coefficients to determine. With very few exceptions, this issue has limited the applications of image-patch methods to the local kind of tasks such as denoising, inpainting, cartoon-texture decomposition, super-resolution, and image deblurring, for which one can process a few patches at a time. Our focus is global imaging tasks such as compressive sensing and medical image recovery, where the whole image is encoded together, making it either impossible or very ineffective to update a few patches at a time. Our strategy is to divide the sparse recovery into multiple subproblems, each of which handles a subset of non-overlapping patches, and then the results of the subproblems are averaged to yield the final recovery. This simple strategy is surprisingly effective in terms of both quality and speed. In addition, we accelerate computation of the learned dictionary by applying a recent block proximal-gradient method, which not only has a lower per-iteration complexity but also takes fewer iterations to converge, compared to the current state-of-the-art. We also establish that our algorithm globally converges to a stationary point. Numerical results on synthetic data demonstrate that our algorithm can recover a more faithful dictionary than two state-of-the-art methods. Combining our whole-image recovery and dictionary-learning methods, we numerically simulate image inpainting, compressive sensing recovery, and deblurring. Our recovery is more faithful than those of a total variation method and a method based on overlapping patches.
Recommendations
- A combined dictionary learning and TV model for image restoration with convergence analysis
- Nonnegative tensor patch dictionary approaches for image compression and deblurring applications
- Hessian Schatten-norm and adaptive dictionary for image recovery
- A fast algorithm for learning overcomplete dictionary for sparse representation based on proximal operators
- On single image scale-up using sparse-representations
Cites work
- $rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
- A block coordinate descent method for regularized multiconvex optimization with applications to nonnegative tensor factorization and completion
- Convergence of a block coordinate descent method for nondifferentiable minimization
- Dictionary Learning Algorithms for Sparse Representation
- Image Deblurring and Super-Resolution by Adaptive Sparse Domain Selection and Adaptive Regularization
- Learning circulant sensing kernels
- Multi-frame compression: theory and design.
- On semi- and subanalytic geometry
- Recursive Least Squares Dictionary Learning Algorithm
- Semianalytic and subanalytic sets
- The Łojasiewicz Inequality for Nonsmooth Subanalytic Functions with Applications to Subgradient Dynamical Systems
Cited in
(10)- Geological facies recovery based on weighted \(\ell_1\)-regularization
- Block Bregman majorization minimization with extrapolation
- Smoothing techniques and difference of convex functions algorithms for image reconstructions
- A globally convergent algorithm for nonconvex optimization based on block coordinate update
- Patch-Based Holographic Image Sensing
- Truncated Nuclear Norm Minimization Based Group Sparse Representation for Image Restoration
- A dictionary learning algorithm for compression and reconstruction of streaming data in preset order
- Nonnegative tensor patch dictionary approaches for image compression and deblurring applications
- First- and second-order methods for online convolutional dictionary learning
- scientific article; zbMATH DE number 7585188 (Why is no real title available?)
This page was built for publication: A fast patch-dictionary method for whole image recovery
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q326397)
