Classifier Fusion: Combination Methods For Semantic Indexing in Video Content

  • Rachid Benmokhtar
  • Benoit Huet
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4132)


Classifier combination has been investigated as a new research field to improve recognition reliability by taking into account the complementarity between classifiers, in particular for automatic semantic-based video content indexing and retrieval. Many combination schemes have been proposed in the literature according to the type of information provided by each classifier as well as their training and adaptation abilities. This paper presents an overview of current research in classifier combination and a comparative study of a number of combination methods. A novel training technique called Weighted Ten Folding based on Ten Folding principle is proposed for combining classifier. Experiments are conducted in the framework of the TRECVID 2005 features extraction task that consists in ordering shots with respect to their relevance to a given class. Finally, we show the efficiency of different combination methods.


Gaussian Mixture Model Video Content Combination Method Semantic Concept Video Shot 
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.
    Kuncheva, L., Bezdek, J.C., Duin, R.: Decision templates for multiple classifier fusion: an experiemental comparaison. Pattern Recognition 34, 299–314 (2001)MATHCrossRefGoogle Scholar
  2. 2.
    Rastrigin, L., Erenstein, R.: Method of collective recognition. Energoizdat (1982)Google Scholar
  3. 3.
    Jacobs, R., Jordan, M., Nowlan, S., Hinton, G.: Adaptive mixtures of local experts. Neural Computation 3, 1409–1431 (1991)CrossRefGoogle Scholar
  4. 4.
    Xu, L., Krzyzak, A., Suen, C.: Methods of combining multiple classifiers and their application to hardwriting recognition. IEEE Trans. Sys. Man. Cyber. 22, 418–435 (1992)CrossRefGoogle Scholar
  5. 5.
    Duin, R., Tax, D.: Experiements with classifier combining rules. In: Kittler, J., Roli, F. (eds.) MCS 2000. LNCS, vol. 1857, pp. 16–29. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  6. 6.
    Jain, A., Duin, R., Mao, J.: Combination of weak classifiers. IEEE Trans. on Pattern Analysis and Machine Intelligence 22(1) (2000)Google Scholar
  7. 7.
    Kuncheva, L.: Fuzzy versus nonfuzzy in combining classifiers designed by bossting. IEEE Transactions on fuzzy systems 11(6) (2003)Google Scholar
  8. 8.
    Chou, K., Tu, L., Shyu, I.: Performances analysis of a multiple classifiers system for recognition of totally unconstrained handwritten numerals. In: 4th International Workshop on Frontiers of Handwritten Recognition, pp. 480–487 (1994)Google Scholar
  9. 9.
    Achermann, B., Bunke, H.: Combination of classifiers on the decision level for face recognition. Technical repport of Bern University (1996)Google Scholar
  10. 10.
    Ho, T.: A theory of multiple classifier systems and its application to visual and word recognition. PhD thesis, Phd thesis of New-York University (1992)Google Scholar
  11. 11.
    Cybenko, G.: Approximations by superposition of a sigmoidal function. Mathematics of Control, Signal and Systems 2, 303–314 (1989)MATHCrossRefMathSciNetGoogle Scholar
  12. 12.
    Paalanen, P., Kamarainen, J., Ilonen, J., Kalviainen, H.: Feature representation and discrimination based on gaussian mixture model probability densities. Research Report, Lappeenranta University of Technology (1995)Google Scholar
  13. 13.
    Souvannavong, F., Merialdo, B., Huet, B.: Multi modal classifier fusion for video shot content retrieval. In: Proceedings of WIAMIS (2005)Google Scholar
  14. 14.
    Freud, Y., Schapire, R.: Experiments with a new boosting algorithms. In: Machine Learning: Proceedings of the 13th International Conference (1996)Google Scholar
  15. 15.
    Skurichina, M., Duin, R.: Bagging for linear classifiers. Pattern Recognition 31(7), 909–930 (1998)CrossRefGoogle Scholar
  16. 16.
    Ma, W., Zhang, H.: Benchmarking of image features for content-based image retrieval. In: Thirtysecond Asilomar Conference on Signals, System and Computers, pp. 253–257 (1998)Google Scholar
  17. 17.
    Carson, C., Thomas, M., Belongie, S.: Blobworld: A system for region-based image indexing and retrieval. In: Third international conference on visual information systems (1999)Google Scholar
  18. 18.
    Souvannavong, F., Merialdo, B., Huet, B.: Latent semantic analysis for an effective region based video shot retrieval system. In: Proceedings of ACM MIR (2004)Google Scholar
  19. 19.
    Wang, R., Huang, T.: Fast camera motion analysis from mpeg domain. In: Proceedings of IEEE ICIP, pp. 691–694 (1999)Google Scholar
  20. 20.
    TRECVID, Digital video retrieval at NIST,
  21. 21.
    Chang, S., Hsu, W., Kennedy, L., Xie, L., Yanagawa, A., Zavesky, E., Zhang, D.: Video seach and high level feature extraction. In: Proceedings of Trecvid (2005)Google Scholar
  22. 22.
    Zheng, Y.Y.C.: Run time information fusion in speech recognition. In: Proc. of ICSLP (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Rachid Benmokhtar
    • 1
  • Benoit Huet
    • 1
  1. 1.Département Communications MultimédiasInstitut EurécomSophia-AntipolisFrance

Personalised recommendations