We consider energy minimization problems related to image labeling, partitioning, and grouping, which typically show up at mid-level stages of computer vision systems. A common feature of these problems is their intrinsic combinatorial complexity from an optimization point-of-view. Rather than trying to compute the global minimum - a goal we consider as elusive in these cases - we wish to design optimization approaches which exhibit two relevant properties: First, in each application a solution with guaranteed degree of suboptimality can be computed. Secondly, the computations are based on clearly defined algorithms which do not comprise any (hidden) tuning parameters.
In this paper, we focus on the second property and introduce a novel and general optimization technique to the field of computer vision which amounts to compute a suboptimal solution by just solving a convex optimization problem. As representative examples, we consider two binary quadratic energy functionals related to image labeling and perceptual grouping. Both problems can be considered as instances of a general quadratic functional in binary variables, which is embedded into a higher-dimensional space such that suboptimal solutions can be computed as minima of linear functionals over cones in that space (semidefi-nite programs). Extensive numerical results reveal that, on the average, suboptimal solutions can be computed which yield a gap below 5% with respect to the global optimum in case where this is known.
Global Optimum Linear Functional Convex Optimization Problem Random Signal Perceptual Grouping
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in to check access.
P.B. Chou and C.M. Brown. Multimodal reconstruction and segmentation with markov random fields and hcf optimization. In Proc. DARPA Image Underst. Workshop, pages 214–221, Cambridge, Massachussetts, April 6–8 1988.Google Scholar
F. Heitz, P. Perez, and P. Bouthemy. Multiscale minimization of global energy functions in some visual recovery problems. Comp. Vis. Graph. Image Proc.: IU, 59(1):125–134, 1994.CrossRefGoogle Scholar
C.-h. Wu and P.C. Doerschuk. Cluster expansions for the deterministic computation of bayesian estimators based on markov random fields. IEEE Trans. Patt. Anal. Mach. Intell., 17(3):275–293, 1995.CrossRefGoogle Scholar
Y. Boykov, O. Veksler, and R. Zabih. Markov random fields with efficient approximations. In Proc. IEEE Conf. on Comp. Vision Patt. Recog. (CVPR’98), pages 648–655, Santa Barbara, California, 1998.Google Scholar
Y.G. Leclerc. Constructing simple stable descriptions for image partitioning. Int. J. of Comp. Vision, 3(1):73–102, 1989.CrossRefGoogle Scholar
A. Blake and A. Zisserman. Visual Reconstruction. MIT Press, 1987.Google Scholar
D. Geiger and F. Girosi. Parallel and deterministic algorithms from mrf’s: Surface reconstruction. IEEE Trans. Patt. Anal. Mach. Intell., 13(5):401–412, 1991.CrossRefGoogle Scholar
L. Herault and R. Horaud. Figure-ground discrimination: A combinatorial optimization approach. IEEE Trans. Patt. Anal. Mach. Intell., 15(9):899–914, 1993.CrossRefGoogle Scholar
S. Gold and A. Rangarajan. A graduated assignment algorithm for graph matching. IEEE Trans. Patt. Anal. Mach. Intell., 18(4):377–388, 1996.CrossRefGoogle Scholar
T. Hofmann and J. Buhmann. Pairwise data clustering by deterministic annealing. IEEE Trans. Patt. Anal. Mach. Intell., 19(1):1–14, 1997.CrossRefGoogle Scholar
M. Sato and S. Ishii. Bifurcations in mean-field-theory annealing. Physical Review E, 53(5):5153–5168, 1996.CrossRefGoogle Scholar
E. Aarts and J.K. Lenstra, editors. Local Search in Combinatorial Optimization, Chichester, 1997. Wiley & Sons.Google Scholar
C. Peterson and B. Söderberg. Artificial neural networks. In Aarts and Lenstra , chapter 7.Google Scholar
C. Schnörr. Unique reconstruction of piecewise smooth images by minimizing strictly convex non-quadratic functionals. J. of Math. Imag. Vision, 4:189–198, 1994.CrossRefGoogle Scholar
C. Schnörr. A study of a convex variational diffusion approach for image segmentation and feature extraction. J. of Math. Imag. and Vision, 8(3):271–292, 1998.zbMATHCrossRefGoogle Scholar
D. Mumford and J. Shah. Optimal approximations by piecewise smooth functions and associated variational problems. Comm. Pure Appl. Math., 42:577–685, 1989.zbMATHCrossRefMathSciNetGoogle Scholar
L. Lovász and A. Schrijver. Cones of matrices and set-functions and 0-1 optimization. SI AM J. Optimization, 1(2): 166–190, 1991.zbMATHCrossRefGoogle Scholar
M.X. Goemans and D.P. Williamson. Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming. J. ACM, 42:1115–1145, 1995.zbMATHCrossRefMathSciNetGoogle Scholar
M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proc. IEEE, 76:869–889, 1988.CrossRefGoogle Scholar
G. Winkler. Image Analysis, Random Fields and Dynamic Monte Carlo Methods, volume 27 of Appl. of Mathematics. Springer-Verlag, Heidelberg, 1995.zbMATHGoogle Scholar
S. Sarkar and K.L. Boyer. Perceptual organization in computer vision: A review and a proposal for a classificatory structure. IEEE Tr. Systems, Man, and Cyb., 23(2):382–399, 1993.CrossRefGoogle Scholar
Y. Nesterov and A. Nemirovskii. Interior Point Polynomial Methods in Convex Programming. SIAM, 1994.Google Scholar
S.J. Benson, Y. Ye, and X. Zhang. Mixed linear and semidefinite programming for combinatorial and quadratic optimization. Optimiz. Methods and Software, 11&12:515–544, 1999.MathSciNetCrossRefGoogle Scholar