Partial Optimality via Iterative Pruning for the Potts Model

  • Paul Swoboda
  • Bogdan Savchynskyy
  • Jörg Kappes
  • Christoph Schnörr
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7893)


We propose a novel method to obtain a part of an optimal non-relaxed integral solution for energy minimization problems with Potts interactions, known also as the minimal partition problem. The method empirically outperforms previous approaches likeMQPBO and Kovtun’s method in most of our test instances and especially in hard ones. As a starting point our approach uses the solution of a commonly accepted convex relaxation of the problem. This solution is then iteratively pruned until our criterion for partial optimality is satisfied. Due to its generality our method can employ any solver for the considered relaxed problem.


Potts Model Segmentation Problem Partial Optimality Energy Minimization Problem Optimal Label 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Brainweb: Simulated brain database,
  2. 2.
  3. 3.
    Alahari, K., Kohli, P., Torr, P.H.S.: Reduce, reuse & recycle: Efficiently solving multi-label MRFs. In: CVPR (2008)Google Scholar
  4. 4.
    Alahari, K., Kohli, P., Torr, P.H.S.: Dynamic hybrid algorithms for MAP inference in discrete MRFs. PAMI 32(10), 1846–1857 (2010)CrossRefGoogle Scholar
  5. 5.
    Boros, E., Hammer, P.L.: Pseudo-Boolean optimization. Discrete Applied Mathematics 123(1-3), 155–225 (2002)MathSciNetzbMATHCrossRefGoogle Scholar
  6. 6.
    Boykov, Y., Kolmogorov, V.: Computing geodesics and minimal surfaces via graph cuts. In: ICCV, pp. 26–33. IEEE Computer Society (2003)Google Scholar
  7. 7.
    Chambolle, A., Cremers, D., Pock, T.: A convex approach for computing minimal partitions. Technical report, Centre des Mathematiques Appliquees, Ecole Polytechnique, Palaiseau, Paris, France (2008)Google Scholar
  8. 8.
    Chambolle, A., Pock, T.: A first-order primal-dual algorithm for convex problems with applications to imaging. Journal of Mathematical Imaging and Vision 40(1), 120–145 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Hammer, P.L., Hansen, P., Simeone, B.: Roof duality, complementation and persistency in quadratic 0-1 optimization. Math. Programming 28, 121–155 (1984)MathSciNetzbMATHCrossRefGoogle Scholar
  10. 10.
    Kahl, F., Strandmark, P.: Generalized roof duality. Discrete Applied Mathematics 160(16-17), 2419–2434 (2012)MathSciNetzbMATHCrossRefGoogle Scholar
  11. 11.
    Kappes, J.H., Andres, B., Hamprecht, F.A., Schnörr, C., Nowozin, S., Batra, D., Kim, S., Kausler, B.X., Lellmann, J., Komodakis, N., Rother, C.: A comparative study of modern inference techniques for discrete energy minimization problem. In: CVPR (2013)Google Scholar
  12. 12.
    Kleinberg, J., Tardos, É.: Approximation algorithms for classification problems with pairwise relationships: metric labeling and Markov random fields. J. ACM 49(5), 616–639 (2002)MathSciNetCrossRefGoogle Scholar
  13. 13.
    Kohli, P., Shekhovtsov, A., Rother, C., Kolmogorov, V., Torr, P.: On partial optimality in multi-label MRFs. In: ICML, pp. 480–487 (2008)Google Scholar
  14. 14.
    Kovtun, I.: Partial optimal labeling search for a NP-hard subclass of (max,+) problems. In: Michaelis, B., Krell, G. (eds.) DAGM 2003. LNCS, vol. 2781, pp. 402–409. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  15. 15.
    Lellmann, J., Schnörr, C.: Continuous multiclass labeling approaches and algorithms. SIAM J. Imag. Sci. 4(4), 1049–1096 (2011)zbMATHCrossRefGoogle Scholar
  16. 16.
    Nemhauser, G.L., Trotter, L.E.: Vertex packings: Structural properties and algorithms. Mathematical Programming 8, 232–248 (1975), doi:10.1007/BF01580444MathSciNetzbMATHCrossRefGoogle Scholar
  17. 17.
    Rother, C., Kolmogorov, V., Lempitsky, V.S., Szummer, M.: Optimizing binary MRFs via extended roof duality. In: CVPR (2007)Google Scholar
  18. 18.
    Shekhovtsov, A., Kolmogorov, V., Kohli, P., Hlavac, V., Rother, C., Torr, P.: LP-relaxation of binarized energy minimization. Research Report CTU–CMP–2007–27, Czech Technical University (2008)Google Scholar
  19. 19.
    Sontag, D.: Approximate Inference in Graphical Models using LP Relaxations. PhD thesis, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science (2010)Google Scholar
  20. 20.
    Windheuser, T., Ishikawa, H., Cremers, D.: Generalized Roof Duality for Multi-Label Optimization: Optimal Lower Bounds and Persistency. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012, Part VI. LNCS, vol. 7577, pp. 400–413. Springer, Heidelberg (2012)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Paul Swoboda
    • 1
  • Bogdan Savchynskyy
    • 2
  • Jörg Kappes
    • 1
  • Christoph Schnörr
    • 1
    • 2
  1. 1.Image and Pattern Analysis GroupUniversity of HeidelbergGermany
  2. 2.Heidelberg Collaboratory for Image ProcessingUniversity of HeidelbergGermany

Personalised recommendations