Combinatorial preconditioners and multilevel solvers for problems in computer vision and image processing

Linear systems and eigen-calculations on symmetric diagonally dominant matrices (SDDs) occur ubiquitously in computer vision, computer graphics, and machine learning. In the past decade a multitude of specialized solvers have been developed to tackle restricted instances of SDD systems for a diverse collection of problems including segmentation, gradient inpainting and total variation. In this paper we explain and apply the support theory of graphs, a set of of techniques developed by the computer science theory community, to construct SDD solvers with provable properties. To demonstrate the power of these techniques, we describe an efficient multigrid-like solver which is based on support theory principles. The solver tackles problems in fairly general and arbitrarily weighted topologies not supported by prior solvers. It achieves state of the art empirical results while providing robust guarantees on the speed of convergence. The method is evaluated on a variety of vision applications.

[1]  Berthold K. P. Horn,et al.  Determining lightness from an image , 1974, Comput. Graph. Image Process..

[2]  Yvan Notay,et al.  Analysis of Aggregation-Based Multigrid , 2008, SIAM J. Sci. Comput..

[3]  Jianbo Shi,et al.  Segmentation given partial grouping constraints , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Gary L. Miller,et al.  A linear work, O(n1/6) time, parallel algorithm for solving planar Laplacians , 2007, SODA '07.

[5]  Y. Notay.  An aggregation-based algebraic multigrid method , 2010 .

[6]  Berthold K. P. Horn,et al.  Determining Optical Flow , 1981, Other Conferences.

[7]  A. Brandt.  General highly accurate algebraic coarsening. , 2000 .

[8]  Gary L. Miller,et al.  Graph partitioning into isolated, high conductance clusters: theory, computation and applications to preconditioning , 2008, SPAA '08.

[9]  Dani Lischinski,et al.  Spectral Matting , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Richard Szeliski,et al.  Locally adapted hierarchical basis preconditioning , 2006, SIGGRAPH '06.

[11]  Laurent D. Cohen,et al.  Non-local Regularization of Inverse Problems , 2008, ECCV.

[12]  Shang-Hua Teng,et al.  Electrical flows, laplacian systems, and faster approximation of maximum flow in undirected graphs , 2010, STOC '11.

[13]  Leo Grady,et al.  A Lattice-Preserving Multigrid Method for Solving the Inhomogeneous Poisson Equations Used in Image Analysis , 2008, ECCV.

[14]  Nancy S. Pollard,et al.  Real-time gradient-domain painting , 2008, ACM Trans. Graph..

[15]  Leo Grady,et al.  Random Walks for Image Segmentation , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Anil Joshi.  Topics in optimization and sparse linear systems , 1997 .

[17]  Christopher V. Alvino,et al.  Reformulating and Optimizing the Mumford-Shah Functional on a Graph - A Faster, Lower Energy Solution , 2008, ECCV.

[18]  Michael F. Cohen,et al.  Fourier Analysis of the 2D Screened Poisson Equation for Gradient Domain Problems , 2008, ECCV.

[19]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  Daniel A. Spielman,et al.  Faster approximate lossy generalized flow via interior point algorithms , 2008, STOC.

[21]  Gary L. Miller,et al.  Solving SDD linear systems in time Õ(mlog nlog(1/ε)) , 2011, ArXiv.

[22]  Peter G. Doyle,et al.  Random Walks and Electric Networks: REFERENCES , 1987 .

[23]  Gary L. Miller,et al.  Approaching optimality for solving SDD systems , 2010, ArXiv.

[24]  A. George,et al.  Graph theory and sparse matrix computation , 1993 .

[25]  O. Axelsson.  Iterative solution methods , 1995 .

[26]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[27]  Berthold K. P. Horn.  SHAPE FROM SHADING: A METHOD FOR OBTAINING THE SHAPE OF A SMOOTH OPAQUE OBJECT FROM ONE VIEW , 1970 .

[28]  L. Rudin,et al.  Nonlinear total variation based noise removal algorithms , 1992 .

[29]  K. Stüben.  Algebraic multigrid (AMG): experiences and comparisons , 1983 .

[30]  Sung Yong Shin,et al.  On pixel-based texture synthesis by non-parametric sampling , 2006, Comput. Graph..

[31]  Leo Grady,et al.  A Seeded Image Segmentation Framework Unifying Graph Cuts And Random Walker Which Yields A New Algorithm , 2007, 2007 IEEE 11th International Conference on Computer Vision.

[32]  Ronen Basri,et al.  Segmentation and boundary detection using multiscale intensity measurements , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[33]  G. Miller,et al.  Combinatorial and algebraic tools for optimal multilevel algorithms , 2007 .

[34]  Daniel Cremers,et al.  Efficient Nonlocal Means for Denoising of Textural Patterns , 2008, IEEE Transactions on Image Processing.

[35]  Michael Elad,et al.  Generalizing the Nonlocal-Means to Super-Resolution Reconstruction , 2009, IEEE Transactions on Image Processing.

[36]  Anders P. Eriksson,et al.  Normalized Cuts Revisited: A Reformulation for Segmentation with Linear Grouping Constraints , 2007, ICCV.

[37]  Leo Grady,et al.  Multilabel random walker image segmentation using prior models , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[38]  Jitendra Malik,et al.  Normalized cuts and image segmentation , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[39]  Jean-Michel Morel,et al.  Nonlocal Image and Movie Denoising , 2008, International Journal of Computer Vision.

[40]  Mikhail Belkin,et al.  Laplacian Eigenmaps for Dimensionality Reduction and Data Representation , 2003, Neural Computation.

[41]  J. W. Ruge,et al.  4. Algebraic Multigrid , 1987 .

[42]  Marian Brezina,et al.  Convergence of algebraic multigrid based on smoothed aggregation , 1998, Numerische Mathematik.

[43]  Zoubin Ghahramani,et al.  Combining active learning and semi-supervised learning using Gaussian fields and harmonic functions , 2003, ICML 2003.

[44]  V. E. Henson,et al.  BoomerAMG: a parallel algebraic multigrid solver and preconditioner , 2002 .

[45]  Leo Grady,et al.  Fast approximate Random Walker segmentation using eigenvector precomputation , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[46]  Tony F. Chan,et al.  Image processing and analysis - variational, PDE, wavelet, and stochastic methods , 2005 .

[47]  Sivan Toledo,et al.  Support-Graph Preconditioners , 2005, SIAM J. Matrix Anal. Appl..

[48]  Ann B. Lee,et al.  Geometric diffusions as a tool for harmonic analysis and structure definition of data: diffusion maps. , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[49]  William Bialek,et al.  Statistics of Natural Images: Scaling in the Woods , 1993, NIPS.

[50]  Gary L. Miller,et al.  Graph Partitioning by Spectral Rounding: Applications in Image Segmentation and Clustering , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[51]  Bruce Hendrickson,et al.  Support Theory for Preconditioning , 2003, SIAM J. Matrix Anal. Appl..

[52]  Gary L. Miller,et al.  Automatic Multiple Retinal Layer Segmentation in Spectral Domain OCT Scans via Spectral Rounding , 2008 .

[53]  Michael I. Jordan,et al.  On Spectral Clustering: Analysis and an algorithm , 2001, NIPS.

[54]  Andrew P. Witkin,et al.  Scale-Space Filtering , 1983, IJCAI.

[55]  Guy Gilboa,et al.  Nonlocal Operators with Applications to Image Processing , 2008, Multiscale Model. Simul..