Non-local Regularization of Inverse Problems
Lecture notes in computer sciencePublished 1 January 2008Open access
Gabriel Peyré, Sébastien Bougleux, Laurent D. Cohen
Citations244
SJR quartileQ2
SJR score0.35
SNIP0.55
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A new framework to regularize linear inverse problems using the total variation on non-local graphs to adapt the penalization to the geometry of the underlying function to recover is proposed.
Abstract
International audience
Keywords
Computer ScienceMathematicsEngineering
IEEE Transactions on Information TheoryCompressed sensing
23,147 Citations2006David L. Donoho
It is shown that "most" subspaces in Ropfm are near-optimal, and that convex optimization (Basis Pursuit) is a near-optimal way to extract information derived from these near-optimal subspaces.
Compressed sensing
17,129 Citations2004David L. Donoho
Elsevier eBooksA Wavelet Tour of Signal Processing
16,401 Citations1999Stéphane Mallat
An introduction to a Transient World and an Approximation Tour of Wavelet Packet and Local Cosine Bases.
Physica D Nonlinear PhenomenaNonlinear total variation based noise removal algorithms
15,672 Citations1992L. Rudin, Stanley Osher +1 more
A constrained optimization type of numerical algorithm for removing noise from images is presented, which appears to be state-of-the-art for very noisy images and yielding sharp edges in the image.
Bilateral filtering for gray and color images
7,991 Citations2002Carlo Tomasi, Roberto Manduchi
In contrast with filters that operate on the three bands of a color image separately, a bilateral filter can enforce the perceptual metric underlying the CIE-Lab color space, and smooth colors and preserve edges in a way that is tuned to human perception.
BiometrikaIdeal spatial adaptation by wavelet shrinkage
7,813 Citations1994David L. Donoho, Iain M. Johnstone
A new principle for spatially-adaptive estimation: selective wavelet reconstruction with an oracle inequality is described and a practical spatially adaptive method, RiskShrink, which works by shrinkage of empirical wavelet coefficients is developed.
IEEE Transactions on Information TheoryNear-Optimal Signal Recovery From Random Projections: Universal Encoding Strategies?
6,835 Citations2006Emmanuel J. Candès, Terence Tao
If the objects of interest are sparse in a fixed basis or compressible, then it is possible to reconstruct f to within very high accuracy from a small number of random measurements by solving a simple linear program.
NatureEmergence of simple-cell receptive field properties by learning a sparse code for natural images
5,840 Citations1996Bruno A. Olshausen, David J. Field
It is shown that a learning algorithm that attempts to find sparse linear codes for natural scenes will develop a complete family of localized, oriented, bandpass receptive fields, similar to those found in the primary visual cortex.
IEEE Transactions on Image ProcessingImage Denoising Via Sparse and Redundant Representations Over Learned Dictionaries
5,385 Citations2006Michael Elad, Michal Aharon
This work addresses the image denoising problem, where zero-mean white and homogeneous Gaussian additive noise is to be removed from a given image, and uses the K-SVD algorithm to obtain a dictionary that describes the image content effectively.
Communications on Pure and Applied MathematicsAn iterative thresholding algorithm for linear inverse problems with a sparsity constraint
4,957 Citations2004Ingrid Daubechies, Michel Defrise +1 more
It is proved that replacing the usual quadratic regularizing penalties by weighted 𝓁p‐penalized penalties on the coefficients of such expansions, with 1 ≤ p ≤ 2, still regularizes the problem.
Multiscale Modeling and SimulationA Review of Image Denoising Algorithms, with a New One
4,004 Citations2005Antoni Buades, B. Coll +1 more
A general mathematical and experimental methodology to compare and classify classical image denoising algorithms and a nonlocal means (NL-means) algorithm addressing the preservation of structure in a digital image are defined.
Image inpainting
3,549 Citations2000Marcelo Bertalmı́o, Guillermo Sapiro +2 more
A novel algorithm for digital inpainting of still images that attempts to replicate the basic techniques used by professional restorators, and does not require the user to specify where the novel information comes from.
International Journal of Computer VisionSUSAN—A New Approach to Low Level Image Processing
3,337 Citations1997Stephen M. Smith, J. Michael Brady
This paper describes a new approach to low level image processing; in particular, edge and corner detection and structure preserving noise reduction and the resulting methods are accurate, noise resistant and fast.
Journal of Mathematical Imaging and VisionAffine Invariant Flows in the Beltrami Framework
3,289 Citations2004Mila Nikolova
The geometric Beltrami framework is shown to incorporate and explain some of the known invariant flows e.g. the equi-affine invariant flow for hypersurfaces, and it is demonstrated that the new concepts put forward enable us to suggest new invariants namely the case where the codimension is greater than one.
IEEE Signal Processing MagazineSuper-resolution image reconstruction: a technical overview
3,251 Citations2003Sung Cheol Park, Min‐Gyu Park +1 more
The goal of this article is to introduce the concept of SR algorithms to readers who are unfamiliar with this area and to provide a review for experts to present the technical review of various existing SR methodologies which are often employed.
IEEE Transactions on Image ProcessingRegion Filling and Object Removal by Exemplar-Based Image Inpainting
3,177 Citations2004Antonio Criminisi, Patrick Pérez +1 more
The simultaneous propagation of texture and structure information is achieved by a single, efficient algorithm that combines the advantages of two approaches: exemplar-based texture synthesis and block-based sampling process.
Texture synthesis by non-parametric sampling
3,036 Citations1999Alexei A. Efros, Thomas Leung
A non-parametric method for texture synthesis that aims at preserving as much local structure as possible and produces good results for a wide variety of synthetic and real-world textures.
Multiscale Modeling and SimulationSignal Recovery by Proximal Forward-Backward Splitting
2,599 Citations2005Patrick L. Combettes, Valérie R. Wajs
It is shown that various inverse problems in signal recovery can be formulated as the generic problem of minimizing the sum of two convex functions with certain regularity properties, which makes it possible to derive existence, uniqueness, characterization, and stability results in a unified and standardized fashion for a large class of apparently disparate problems.
Mathematical ProgrammingSmooth minimization of non-smooth functions
2,522 Citations2004Yu. Nesterov
A new approach for constructing efficient schemes for non-smooth convex optimization is proposed, based on a special smoothing technique, which can be applied to functions with explicit max-structure, and can be considered as an alternative to black-box minimization.
IEEE Computer Graphics and ApplicationsExample-based super-resolution
2,509 Citations2002William T. Freeman, Thouis R. Jones +1 more
This work built on another training-based super- resolution algorithm and developed a faster and simpler algorithm for one-pass super-resolution that requires only a nearest-neighbor search in the training set for a vector derived from each patch of local image data.
Proceedings of the National Academy of SciencesGeometric diffusions as a tool for harmonic analysis and structure definition of data: Diffusion maps
1,764 Citations2005Ronald R. Coifman, Stéphane Lafon +5 more
The process of iterating or diffusing the Markov matrix is seen as a generalization of some aspects of the Newtonian paradigm, in which local infinitesimal transitions of a system lead to global macroscopic descriptions by integration.
IEEE Transactions on Image ProcessingSparse Representation for Color Image Restoration
1,721 Citations2007Julien Mairal, Michael Elad +1 more
This work puts forward ways for handling nonhomogeneous noise and missing information, paving the way to state-of-the-art results in applications such as color image denoising, demosaicing, and inpainting, as demonstrated in this paper.
Multiscale Modeling and SimulationNonlocal Operators with Applications to Image Processing
1,415 Citations2008Guy Gilboa, Stanley Osher
This topic can be viewed as an extension of spectral graph theory and the diffusion geometry framework to functional analysis and PDE-like evolutions to define new types of flows and functionals for image processing and elsewhere.
Fast texture synthesis using tree-structured vector quantization
1,347 Citations2000Li‐Yi Wei, Marc Levoy
This paper presents an efficient algorithm for realistic texture synthesis derived from Markov Random Field texture models and generates textures through a deterministic searching process that accelerates this synthesis process using tree-structured vector quantization.
SIAM Journal on Applied MathematicsMathematical Models for Local Nontexture Inpaintings
1,175 Citations2002Jianhong Shen, Tony F. Chan
The broad applications of the inpainting models are demonstrated through restoring scratched old photos, disocclusion in vision analysis, text removal, digital zooming, and edge-based image coding.
IEEE Transactions on Image ProcessingFilling-in by joint interpolation of vector fields and gray levels
967 Citations2001Coloma Ballester, Marcelo Bertalmı́o +3 more
A variational approach for filling-in regions ofMissing data in digital images is introduced, based on joint interpolation of the image gray levels and gradient/isophotes directions, smoothly extending in an automatic fashion the isophote lines into the holes of missing data.
Applied and Computational Harmonic AnalysisSimultaneous cartoon and texture image inpainting using morphological component analysis (MCA)
906 Citations2005Michael Elad, Jean‐Luc Starck +2 more
A novel inpainting algorithm that is capable of filling in holes in overlapping texture and cartoon image layers using a direct extension of a recently developed sparse-representation-based image decomposition method called MCA (morphological component analysis).
SIAM Journal on Imaging SciencesBregmanized Nonlocal Regularization for Deconvolution and Sparse Reconstruction
694 Citations2010Xiaoqun Zhang, Martin Burger +2 more
The proposed general algorithm framework for inverse problem regularization with a single forward-backward operator step, namely, Bregmanized operator splitting (BOS), converges without fully solving the subproblems, and numerical results on deconvolution and compressive sensing illustrate the performance of nonlocal total variation regularization under the proposed algorithm framework.
IEEE Transactions on Pattern Analysis and Machine IntelligenceVector-valued image regularization with PDEs: a common framework for different applications
670 Citations2005David Tschumperlé, Rachid Deriche
A unifying expression is proposed that gathers the majority of PDE-based formalisms for vector-valued image regularization into a single generic anisotropic diffusion equation, allowing us to implement the authors' regularization framework with accuracy by taking the local filtering properties of the proposed equations into account.
International Journal of Imaging Systems and TechnologyAdvances and challenges in super‐resolution
617 Citations2004Sina Farsiu, D. Robinson +2 more
A detailed study of several very important aspects of Super‐Resolution, often ignored in the literature, are presented, and robustness, treatment of color, and dynamic operation modes are discussed.
Journal of Mathematical Imaging and VisionImage Sharpening by Flows Based on Triple Well Potentials
616 Citations2004Guy Gilboa, Nir Sochen +1 more
Image sharpening in the presence of noise is formulated as a non-convex variational problem that incorporates a gradient-dependent potential, a convex fidelity criterion and a high order convex regularizing term.
Multiscale Modeling and SimulationNonlocal Linear Image Regularization and Supervised Segmentation
456 Citations2007Guy Gilboa, Stanley Osher
The steepest descent for minimizing the functional is interpreted as a nonlocal diffusion process, which allows a convenient framework for nonlocal variational minimizations, including variational denoising, Bregman iterations, and the recently proposed inverse scale space.
IEEE Transactions on Image ProcessingNonlocal Discrete Regularization on Weighted Graphs: A Framework for Image and Manifold Processing
455 Citations2008Abderrahim Elmoataz, Olivier Lézoray +1 more
A nonlocal discrete regularization framework on weighted graphs of the arbitrary topologies for image and manifold processing, which leads to a family of simple and fast nonlinear processing methods based on the weighted -Laplace operator, parameterized by the degree of regularity, the graph structure and the graph weight function.
Multiscale Modeling and SimulationDeblurring and Denoising of Images by Nonlocal Functionals
364 Citations2005Stefan Kindermann, Stanley Osher +1 more
A nonlocal variant of the well-known bounded variation regularization, which does not suffer from the staircase effect is defined and existence of a minimizer of the corresponding regularization functional for the denoising and deblurring problem is shown.
The Computer JournalInpainting and Zooming Using Sparse Representations
332 Citations2007M.J. Fadili, Jean‐Luc Starck +1 more
An easy and efficient sparse-representation-based iterative algorithm for image inpainting that allows a high degree of flexibility to recover different structural components in the image (piecewise smooth, curvilinear, texture, etc.).
On image denoising methods
212 Citations2004Antoni Buades, B. Coll +1 more
A general mathematical and experimental methodology to compare and classify classical image denoising algorithms is defined, and an algorithm (Non Local Means) addressing the preservation of structure in a digital image is proposed.
Multiscale Modeling and SimulationBandelet Image Approximation and Compression
211 Citations2005Erwan Le Pennec, Stéphane Mallat
For functions that are uniformly regular outside a set of edge curves that are geometrically regular, the main theorem proves that bandelet approximations satisfy an optimal asymptotic error decay rate.
IEEE Transactions on Image ProcessingDisocclusion: a variational approach using level lines
184 Citations2002Simon Masnou
It is shown in this paper how disocclusion can be performed by means of the level-lines structure, which offers a reliable, complete and contrast-invariant representation of images.
IEEE Transactions on ComputersComputation of the Fast Walsh-Fourier Transform
179 Citations1969John L. Shanks
An efficient Walsh transform computation algorithm is derived which is analogous to the Cooley-Tukey algorithm for the complex-exponential Fourier transform.
Journal of Mathematical Imaging and VisionSparse Modeling of Textures
158 Citations2008Gabriel Peyré
A generative model for textures that uses a local sparse description of the image content that enforces the sparsity of the expansion of local texture patches on adapted atomic elements is presented.
SIAM Journal on Numerical AnalysisEdge Direction Preserving Image Zooming: A Mathematical and Numerical Analysis
154 Citations2001François Malgouyres, Frédéric Guichard
This paper focuses on linear methods, gives a general framework to design them, and shows that the preservation of 1D structures pleads in favor of the cancellation of the periodization of the image spectrum, and studies variational nonlinear methods.
Lecture notes in computer scienceRegularization on Discrete Spaces
149 Citations2005Dengyong Zhou, Bernhard Schölkopf
A general discrete regularization framework defined on finite object sets is proposed, which can be thought of as discrete analogue of classical regularization theory, built on discrete analysis and geometry developed by ourselves.
Journal of Mathematical Imaging and VisionSome First-Order Algorithms for Total Variation Based Image Restoration
146 Citations2009Jean–François Aujol
An old and forgotten algorithm that is revealed wider than recent schemes and able to improve contemporary schemes is revealed and proved to be optimal among first-order numerical schemes for image restoration.
Lecture notes in computer scienceA l 1-Unified Variational Framework for Image Restoration
145 Citations2004Julien Bect, Laure Blanc‐Féraud +2 more
Among image restoration literature, there are mainly two kinds of approach, one is based on a process over image wavelet coefficients, as wavelet shrinkage for denoising and the other is based over image gradient, which usually assumes that the image belongs to the space of functions of Bounded Variation.
SIAM Journal on Imaging SciencesLearning the Morphological Diversity
96 Citations2010Gabriel Peyré, Jalal Fadili +1 more
A new method for image separation into a linear combination of morphological components is proposed, using adapted dictionaries learned from data to circumvent some difficulties faced by fixed dictionaries.
SIAM Journal on Mathematical AnalysisExemplar-Based Inpainting from a Variational Point of View
90 Citations2010Jean–François Aujol, Saïd Ladjal +1 more
The purpose in this paper is to propose well-posed variational models in the continuous domain that can be naturally associated to exemplar-based algorithms and to investigate their ability to reconsider geometry.
Multiscale Modeling and SimulationImage Processing with Nonlocal Spectral Bases
87 Citations2008Gabriel Peyré
Numerical results compare the efficiency of PDE flows, energy minimizations and thresholding operators in the semi-local and non-local settings as well as the performance of non-linear approximation in orthogonal bases.
Multidimensional Systems and Signal ProcessingExample-based single document image super-resolution: a global MAP approach with outlier rejection
87 Citations2007D. Datsenko, Michael Elad
This paper proposes an efficient scheme for using image examples as driving a powerful regularization, applied to the image scale-up (super-resolution) problem, and demonstrates the algorithm on several scanned documents with promising results.
IEEE Transactions on Image ProcessingA Short- Time Beltrami Kernel for Smoothing Images and Manifolds
81 Citations2007Alon Spira, Ron Kimmel +1 more
A short-time kernel is introduced for the Beltrami image enhancing flow that combines the geometry of the image and that of the manifold into one metric tensor, thus enabling a natural unified approach for the manipulation of both.
Lecture notes in computer scienceSolving the Inverse Problem of Image Zooming Using “Self-Examples”
78 Citations2007Mehran Ebrahimi, Edward R. Vrscay
A novel single-frame image zooming technique based on so-called "self-examples" that targets the MMSE estimate by learning the posterior directly from examples taken from the image itself at a different scale, similar to fractal-based techniques.
Pattern Recognition LettersA non-local regularization strategy for image deconvolution
74 Citations2008Max Mignotte
IEEE Transactions on Pattern Analysis and Machine IntelligenceTexture Synthesis with Grouplets
59 Citations2009Gabriel Peyré
Lecture notes in computer scienceExemplar-Based Interpolation of Sparsely Sampled Images
38 Citations2009Gabriele Facciolo, Pablo Arias +2 more
Initial experimental results with the proposed framework, at very low sampling densities, are very encouraging, and some departures from the variational setting are explored, showing a remarkable ability to recover textures at low sampling density.
Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIENonlocal evolutions for image regularization
33 Citations2007Guy Gilboa, Stanley Osher
A nonlocal quadratic functional of weighted differences, which represents the affinity between different pixels in the image, and shows state of the art denoising results using the nonlocal flow.
Some algorithms for total variation based image restoration
25 Citations2008Jean–François Aujol
The convergence of all the proposed schemes for image restoration rely on a duality-based algorithm proposed in 1979 by Bermudez and Moreno, and on general minimization schemes recently developped by Y. Nesterov.