Convolutional proximal neural networks and plug-and-play algorithms

DOI10.1016/j.laa.2021.09.004OpenAlexW3196943155MaRDI QIDQ2238870

Johannes Hertrich, Sebastian Neumayer, Gabriele Drauschke

Publication date: 2 November 2021

Published in: Linear Algebra and its Applications (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2011.02281

zbMATH Keywords

neural networksdenoisingmatrix manifoldsaveraged operatorsLipschitz networksplug-and-play algorithms

Mathematics Subject Classification ID

Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Applications of functional analysis in optimization, convex analysis, mathematical programming, economics (46N10) Inverse problems in optimal control (49N45) Neural nets and related approaches to inference from stochastic processes (62M45) Computer science (68-XX) Information and communication theory, circuits (94-XX)

Related Items

Speckle reduction in matrix-log domain for synthetic aperture radar imaging ⋮ WARPd: A Linearly Convergent First-Order Primal-Dual Algorithm for Inverse Problems with Approximate Sharpness Conditions ⋮ Inertial stochastic PALM and applications in machine learning ⋮ Image Denoising: The Deep Learning Revolution and Beyond—A Survey Paper ⋮ Approximation of Lipschitz Functions Using Deep Spline Neural Networks ⋮ Self-Supervised Deep Learning for Image Reconstruction: A Langevin Monte Carlo Approach ⋮ WPPNets and WPPFlows: The Power of Wasserstein Patch Priors for Superresolution ⋮ Convergence Results for Primal-Dual Algorithms in the Presence of Adjoint Mismatch ⋮ Dynamical Systems–Based Neural Networks ⋮ On \(\alpha\)-firmly nonexpansive operators in \(r\)-uniformly convex spaces

Uses Software

• UNLocBoX
• BSDS
• Adam
• mftoolbox
• BM3D
• DnCNN

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• A feasible method for optimization with orthogonality constraints
• Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers
• Nonlinear total variation based noise removal algorithms
• The Little Engine that Could: Regularization by Denoising (RED)
• Operator splittings, Bregman methods and frame shrinkage in image processing
• Compositions and convex combinations of averaged nonexpansive operators
• Parseval proximal neural networks
• Spectral and computational properties of band symmetric Toeplitz matrices
• On the Douglas-Rachford splitting method and the proximal point algorithm for maximal monotone operators
• Preconditioners for ill-conditioned Toeplitz matrices
• Monotone operator theory in convex optimization
• Variational networks: an optimal control approach to early stopping variational methods for image restoration
• Deep neural network structures solving variational inequalities
• Splitting Methods in Communication, Imaging, Science, and Engineering
• Proximal Splitting Methods in Signal Processing
• Functions of Difference Matrices Are Toeplitz Plus Hankel
• Estimation of the Noise Level Function Based on a Nonparametric Detection of Homogeneous Image Regions
• Proximal Thresholding Algorithm for Minimization over Orthonormal Bases
• A New Scaling for Newton's Iteration for the Polar Decomposition and its Backward Stability
• Computing the Polar Decomposition—with Applications
• First-Order Methods in Optimization
• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• An iterative thresholding algorithm for linear inverse problems with a sparsity constraint
• Solving monotone inclusions via compositions of nonexpansive averaged operators
• Stabilizing Invertible Neural Networks Using Mixture Models
• Lipschitz Certificates for Layered Network Structures Driven by Averaged Activation Operators
• Quantitative Convergence Analysis of Iterated Expansive, Set-Valued Mappings
• First Order Algorithms in Variational Image Processing
• BM3D Frames and Variational Image Deblurring
• Functions of Matrices
• Signal Recovery by Proximal Forward-Backward Splitting
• Proximité et dualité dans un espace hilbertien
• Mean Value Methods in Iteration