Image Labeling Based on Graphical Models Using Wasserstein Messages and Geometric Assignment

DOI10.1137/17M1150669zbMath1401.62099arXiv1710.01493OpenAlexW3103059853MaRDI QIDQ4689763

Christoph Schnörr, Ruben Hühnerbein, Fabrizio Savarino, Freddie Åström

Publication date: 17 October 2018

Published in: SIAM Journal on Imaging Sciences (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1710.01493

zbMATH Keywords

graphical modelsWasserstein distancediscrete optimal transportFisher-Rao metricentropic regularizationimage labelingassignment manifoldRiemannian gradient flow

Mathematics Subject Classification ID

Random fields; image analysis (62M40) Image analysis in multivariate analysis (62H35) Numerical optimization and variational techniques (65K10) Computing methodologies for image processing (68U10)

Related Items

Assignment Flows, Learning adaptive regularization for image labeling using geometric assignment, Geometric numerical integration of the assignment flow

Uses Software

• MAXFLOW
• OpenGM

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• A sparse multiscale algorithm for dense optimal transport
• Matrix scaling, entropy minimization, and conjugate duality. II: The dual problem
• Variational problems in SBV and image segmentation
• Nonlocal discrete \(p\)-Laplacian driven image and manifold processing
• The Boolean quadratic polytope: Some characteristics, facets and relatives
• On a special class of convex functions
• Proximal minimization algorithm with \(D\)-functions
• Interior point methods of mathematical programming
• Convex analysis approach to d. c. programming: Theory, algorithms and applications
• Image labeling by assignment
• Linear programming, complexity theory and elementary functional analysis
• Iterative multiplicative filters for data labeling
• A Smoothed Dual Approach for Variational Wasserstein Problems
• Nonlocal Operators with Applications to Image Processing
• Entropic Approximation of Wasserstein Gradient Flows
• Approximation algorithms for classification problems with pairwise relationships
• MAP Estimation Via Agreement on Trees: Message-Passing and Linear Programming
• Constructing Free-Energy Approximations and Generalized Belief Propagation Algorithms
• Graphical Models, Exponential Families, and Variational Inference
• The Sinkhorn–Knopp Algorithm: Convergence and Applications
• Riemannian center of mass and mollifier smoothing
• Monotone Operators and the Proximal Point Algorithm
• Variational Analysis
• A D.C. Optimization Algorithm for Solving the Trust-Region Subproblem
• A Graph Framework for Manifold-Valued Data
• Diffuse Interface Models on Graphs for Classification of High Dimensional Data
• Norm-Product Belief Propagation: Primal-Dual Message-Passing for Approximate Inference
• A Linear Programming Formulation and Approximation Algorithms for the Metric Labeling Problem
• A Relationship Between Arbitrary Positive Matrices and Doubly Stochastic Matrices
• The Theory of Max-Min, with Applications
• Estimation and Marginalization Using the Kikuchi Approximation Methods
• Numerical Integration of Riemannian Gradient Flows for Image Labeling
• MAP Image Labeling Using Wasserstein Messages and Geometric Assignment
• Geometry of cuts and metrics