Learning a Loopy Model Exactly

Andreas Christian Müller, Sven Behnke

Abstract

Learning structured models using maximum margin techniques has become an indispensable tool for computer vision researchers, as many computer vision applications can be cast naturally as an image labeling problem. Pixel-based or superpixel-based conditional random fields are particularly popular examples. Typically, neighborhood graphs, which contain a large number of cycles, are used. As exact inference in loopy graphs is NP-hard in general, learning these models without approximations is usually deemed infeasible. In this work we show that, despite the theoretical hardness, it is possible to learn loopy models exactly in practical applications. To this end, we analyze the use of multiple approximate inference techniques together with cutting plane training of structural SVMs. We show that our proposed method yields exact solutions with an optimality guarantees in a computer vision application, for little additional computational cost. We also propose a dynamic caching scheme to accelerate training further, yielding runtimes that are comparable with approximate methods. We hope that this insight can lead to a reconsideration of the tractability of loopy models in computer vision.

References

  1. Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., and S üsstrunk, S. (2012). SLIC Superpixels Compared to State-of-the-Art Superpixel Methods. Pattern Analysis and Machine Intelligence.
  2. Dahl, J. and Vandenberghe, L. (2006). Cvxopt: A python package for convex optimization. In European Converence on Computer Vision.
  3. Dann, C., Gehler, P., Roth, S., and Nowozin, S. (2012). Pottics-the potts topic model for semantic image segmentation. In German Conference on Pattern Recognition (DAGM).
  4. Everingham, M., Van Gool, L., Williams, C. K. I., Winn, J., and Zisserman, A. (2010). The Pascal Visual Object Classes (VOC) Challenge. International Journal of Computer Vision, 88.
  5. Finley, T. and Joachims, T. (2008). Training structural SVMs when exact inference is intractable. In International Conference on Machine Learning.
  6. Fulkerson, B., Vedaldi, A., and Soatto, S. (2009). Class segmentation and object localization with superpixel neighborhoods. In International Converence on Computer Vision.
  7. Gonfaus, J. M., Boix, X., van de Weijer, J., Bagdanov, A. D., Serrat, J., and Gonzalez, J. (2010). Harmony potentials for joint classification and segmentation. In Computer Vision and Pattern Recognition.
  8. Hazan, T. and Urtasun, R. (2010). A primal-dual messagepassing algorithm for approximated large scale structured prediction. In Neural Information Processing Systems.
  9. Joachims, T., Finley, T., and Yu, C.-N. J. (2009). Cuttingplane training of structural SVMs. Machine Learning, 77(1).
  10. Kohli, P., Torr, P. H., et al. (2009). Robust higher order potentials for enforcing label consistency. International Journal of Computer Vision, 82(3).
  11. Komodakis, N. (2011). Efficient training for pairwise or higher order crfs via dual decomposition. In Computer Vision and Pattern Recognition.
  12. Krähenbühl, P. and Koltun, V. (2012). Efficient inference in fully connected CRFs with Gaussian edge potentials.
  13. Krähenbühl, P. and Koltun, V. (2013). Parameter learning and convergent inference for dense random fields. In International Conference on Machine Learning.
  14. Ladicky, L., Russell, C., Kohli, P., and Torr, P. H. (2009). Associative hierarchical CRFs for object class image segmentation. In International Converence on Computer Vision.
  15. Lempitsky, V., Rother, C., Roth, S., and Blake, A. (2010). Fusion moves for markov random field optimization. Pattern Analysis and Machine Intelligence, 32(8).
  16. Lucchi, A., Li, Y., Boix, X., Smith, K., and Fua, P. (2011). Are spatial and global constraints really necessary for segmentation? In International Converence on Computer Vision.
  17. Lucchi, A., Li, Y., and Fua, P. (2013). Learning for structured prediction using approximate subgradient descent with working sets. In Computer Vision and Pattern Recognition.
  18. Martins, A. F., Figueiredo, M. A., Aguiar, P. M., Smith, N. A., and Xing, E. P. (2011). An augmented lagrangian approach to constrained map inference. In International Conference on Machine Learning.
  19. Meshi, O., Sontag, D., Jaakkola, T., and Globerson, A. (2010). Learning efficiently with approximate inference via dual losses. In International Conference on Machine Learning.
  20. Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., et al. (2011). Scikit-learn: Machine learning in python. Journal of Machine Learning Research, 12.
  21. Rother, C., Kolmogorov, V., Lempitsky, V., and Szummer, M. (2007). Optimizing binary MRFs via extended roof duality. In Computer Vision and Pattern Recognition.
  22. Szummer, M., Kohli, P., and Hoiem, D. (2008). Learning CRFs using graph cuts. In European Converence on Computer Vision.
  23. Taskar, B., Guestrin, C., and Koller, D. (2003). Maxmargin markov networks. Neural Information Processing Systems.
  24. Tsochantaridis, I., Joachims, T., Hofmann, T., Altun, Y., and Singer, Y. (2006). Large margin methods for structured and interdependent output variables. Journal of Machine Learning Research, 6(2).
  25. Vedaldi, A. and Fulkerson, B. (2008). VLFeat: An open and portable library of computer vision algorithms. http://www.vlfeat.org/.
  26. Vedaldi, A. and Zisserman, A. (2010). Efficient additive kernels via explicit feature maps. In Computer Vision and Pattern Recognition.
  27. Xia, W., Song, Z., Feng, J., Cheong, L.-F., and Yan, S. (2012). Segmentation over detection by coupled global and local sparse representations. In European Converence on Computer Vision.
Download


Paper Citation


in Harvard Style

Christian Müller A. and Behnke S. (2014). Learning a Loopy Model Exactly . In Proceedings of the 9th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2014) ISBN 978-989-758-004-8, pages 337-344. DOI: 10.5220/0004674503370344


in Bibtex Style

@conference{visapp14,
author={Andreas Christian Müller and Sven Behnke},
title={Learning a Loopy Model Exactly},
booktitle={Proceedings of the 9th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2014)},
year={2014},
pages={337-344},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004674503370344},
isbn={978-989-758-004-8},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 9th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2014)
TI - Learning a Loopy Model Exactly
SN - 978-989-758-004-8
AU - Christian Müller A.
AU - Behnke S.
PY - 2014
SP - 337
EP - 344
DO - 10.5220/0004674503370344