Novel Fusion Methods for Pattern Recognition

  • Muhammad Awais
  • Fei Yan
  • Krystian Mikolajczyk
  • Josef Kittler
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6911)


Over the last few years, several approaches have been proposed for information fusion including different variants of classifier level fusion (ensemble methods), stacking and multiple kernel learning (MKL). MKL has become a preferred choice for information fusion in object recognition. However, in the case of highly discriminative and complementary feature channels, it does not significantly improve upon its trivial baseline which averages the kernels. Alternative ways are stacking and classifier level fusion (CLF) which rely on a two phase approach. There is a significant amount of work on linear programming formulations of ensemble methods particularly in the case of binary classification.

In this paper we propose a multiclass extension of binary ν-LPBoost, which learns the contribution of each class in each feature channel. The existing approaches of classifier fusion promote sparse features combinations, due to regularization based on ℓ1-norm, and lead to a selection of a subset of feature channels, which is not good in the case of informative channels. Therefore, we generalize existing classifier fusion formulations to arbitrary ℓ p -norm for binary and multiclass problems which results in more effective use of complementary information. We also extended stacking for both binary and multiclass datasets. We present an extensive evaluation of the fusion methods on four datasets involving kernels that are all informative and achieve state-of-the-art results on all of them.


Support Vector Machine Fusion Method Ensemble Method Mean Average Precision Fusion Scheme 
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.
    Bach, F., Lanckriet, G., Jordan, M.: Multiple Kernel Learning, Conic Duality, and the SMO Algorithm. In: ICML (2004)Google Scholar
  2. 2.
    Džeroski, S., Ženko, B.: Is combining classifiers with stacking better than selecting the best one? ML 54(3), 255–273 (2004)zbMATHGoogle Scholar
  3. 3.
    Everingham, M., Van Gool, L., Williams, C.K.I., Winn, J., Zisserman, A.: The pascal visual object classes (voc) challenge. IJCV 88(2), 303–338 (2010)CrossRefGoogle Scholar
  4. 4.
    Fei-Fei, L., Fergus, R., Perona, P.: One-shot Learning of Object Categories. PAMI, 594–611 (2006)Google Scholar
  5. 5.
    Freund, Y., Schapire, R.: A Desicion-Theoretic Generalization of On-Line Learning and an Application to Boosting. In: CLT (1995)Google Scholar
  6. 6.
    Gehler, P., Nowozin, S.: On Feature Combination for Multiclass Object Classification. In: ICCV (2009)Google Scholar
  7. 7.
    Griffin, G., Holub, A., Perona, P.: Caltech-256 object category dataset. Tech. Rep. 7694, California Institute of Technology (2007)Google Scholar
  8. 8.
    Kittler, J., Hatef, M., Duin, R., Matas, J.: On combining classifiers. PAMI 20(3), 226–239 (1998)CrossRefGoogle Scholar
  9. 9.
    Kloft, M., Brefeld, U., Sonnenburg, S., Zien, A., Laskov, P., Müller, K.: Efficient and Accurate lp-norm MKL. In: NIPS (2009)Google Scholar
  10. 10.
    Lanckriet, G., Cristianini, N., Bartlett, P., Ghaoui, L., Jordan, M.: Learning the Kernel Matrix with Semidefinite Programming. JMLR 5, 27–72 (2004)MathSciNetzbMATHGoogle Scholar
  11. 11.
    Lazebnik, S., Schmid, C., Ponce, J.: Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories. In: CVPR (2006)Google Scholar
  12. 12.
    Mikolajczyk, K., Schmid, C.: A performance evaluation of local descriptors. PAMI 27(10), 1615–1630 (2005)CrossRefGoogle Scholar
  13. 13.
    Nilsback, M.E., Zisserman, A.: Automated Flower Classification over a Large Number of Classes. In: ICCVGIP (2008)Google Scholar
  14. 14.
    Nilsback, M., Zisserman, A.: A visual Vocabulary for Flower Classification. In: CVPR (2006)Google Scholar
  15. 15.
    Rakotomamonjy, A., Bach, F., Canu, S., Grandvalet, Y.: SimpleMKL. JMLR 9, 2491–2521 (2008)MathSciNetzbMATHGoogle Scholar
  16. 16.
    Rätsch, G., Schölkopf, B., Smola, A., Mika, S., Müller, K., Onoda, T.: Robust Ensemble Learning for Data Analysis. In: PACKDDM (2000)Google Scholar
  17. 17.
    Read, J., Pfahringer, B., Holmes, G., Frank, E.: Classifier chains for multi-label classification. In: MLKDD, pp. 254–269 (2009)Google Scholar
  18. 18.
    Rifkin, R., Klautau, A.: In defense of one-vs-all classification. JMLR 5, 101–141 (2004)MathSciNetzbMATHGoogle Scholar
  19. 19.
    van de Sande, K., Gevers, T., Snoek, C.: Evaluation of color descriptors for object and scene recognition. In: CVPR (2008)Google Scholar
  20. 20.
    Sonnenburg, S., Rätsch, G., Schafer, C., Schölkopf, B.: Large Scale Multiple Kernel Learning. JMLR 7, 1531–1565 (2006)MathSciNetzbMATHGoogle Scholar
  21. 21.
    Wolpert, D.: Stacked generalization. Neural Networks 5(2), 241–259 (1992)CrossRefGoogle Scholar
  22. 22.
    Xie, N., Ling, H., Hu, W., Zhang, Z.: Use bin-ratio information for category and scene classification. In: CVPR (2010)Google Scholar
  23. 23.
    Yan, F., Mikolajczyk, K., Barnard, M., Cai, H., Kittler, J.: Lp norm multiple kernel fisher discriminant analysis for object and image categorisation. In: CVPR (2010)Google Scholar
  24. 24.
    Ying, Y., Huang, K., Campbell, C.: Enhanced protein fold recognition through a novel data integration approach. BMCB 10(1), 267 (2009)Google Scholar
  25. 25.
    Zien, A., Ong, C.: Multiclass Multiple Kernel Learning. In: ICML (2007)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Muhammad Awais
    • 1
  • Fei Yan
    • 1
  • Krystian Mikolajczyk
    • 1
  • Josef Kittler
    • 1
  1. 1.Centre for Vision, Speech and Signal Processing (CVSSP)University of SurreyUK

Personalised recommendations