We address the problem of combining different types of audio features for music classification. Several feature-level and decision-level combination methods have been studied, including kernel methods based on multiple kernel learning, decision level fusion rules and stacked generalization. Eight widely used audio features were examined in the experiments on multi-feature based music classification. Results on benchmark data set have demonstrated the effectiveness of using multiple types of features for music classification and identified the most effective combination method for improving classification performance.


Feature Vector Combination Method Fusion Rule Feature Combination Linear Support Vector Machine 
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.


  1. 1.
    Varma, M., Ray, D.: Learning the discriminative power invariance trade-off. In: Intl. Conf. on Computer Vision (2007)Google Scholar
  2. 2.
    Lanckriet, G., Cristianini, N., Bartlett, P., Ghaoui, L.E., Jordan, M.I.: Learning the kernel matrix with semidefinite programming. Journal of Machine Learning Research 5, 27–72 (2004)Google Scholar
  3. 3.
    Kittler, J., Hatef, M., Duin, R., Matas, J.: On combining classifiers. IEEE Transactions on Pattern Analysis and Machine Intelligence 20(3), 226–239 (1998)CrossRefGoogle Scholar
  4. 4.
    Wolpert, D.: Stacked generalization. Neural Networks 5(2), 241–259 (1992)CrossRefGoogle Scholar
  5. 5.
    Ting, K.M., Witten, I.: Issues in stacked generalization. Journal of Artificial Intelligence Research 10, 271–289 (1999)zbMATHGoogle Scholar
  6. 6.
    Boser, B.E., Guyon, I., Vapnik, V.: A training algorithm for optimal margin classifiers. In: ACM Conf. Computational Learning Theory, pp. 144–152 (1992)Google Scholar
  7. 7.
    Mandel, M., Ellis, D.: Song-level features and svms for music classification. In: Intl. Conf. Music Information Retrieval (2005)Google Scholar
  8. 8.
    Pampalk, E., Rauber, A., Merkl, D.: Content-based organization and visualization of music archives. In: ACM Multimedia (2002)Google Scholar
  9. 9.
    Tzanetakis, G., Cook, P.: Musical genre classification of audio signals. IEEE Trans. Speech and Audio Processing 10(5), 293–302 (2002)CrossRefGoogle Scholar
  10. 10.
    Bergstra, J., Casagrande, N., Erhan, D., Eck, D., Kegl, B.: Aggregate features and ada boost for music classification. Machine Learning 65(2-3), 473–484 (2006)CrossRefGoogle Scholar
  11. 11.
    Lee, C.H., Shih, J.L., Yu, K.M., Lin, H.S.: Automatic music genre classification based on modulation spectral analysis of spectral and cepstral features. IEEE Trans. Multimedia 11(4), 670–682 (2009)CrossRefGoogle Scholar
  12. 12.
    Kim, H.G., Moreau, N., Sikora, T.: Audio classification based on mpeg-7 spectral basis representation. IEEE Trans. Circuits and Systems for Video Technology 14(5), 716–725 (2004)CrossRefGoogle Scholar
  13. 13.
    Lu, L., Liu, D., Zhang, H.J.: Automatic mood detection and tracking of music audio signals. IEEE Trans. Speech and Audio Processing 14(1), 5–18 (2006)CrossRefMathSciNetGoogle Scholar
  14. 14.
    Cheng, H.T., Yang, Y.H., Lin, Y.C., Liao, I.B., Chen, H.H.: Automatic chord recognition for music classification and retrieval. In: Intl. Conf. Multi. Expo. (2008)Google Scholar
  15. 15.
    Guyon, I., Weston, J., Barnhill, S., Vapnik, V.: Gene selection for cancer classification using support vector machines. Machine Learning 46(1-3), 389–422 (2002)zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Zhouyu Fu
    • 1
  • Guojun Lu
    • 1
  • Kai-Ming Ting
    • 1
  • Dengsheng Zhang
    • 1
  1. 1.Gippsland School of ITMonash UniversityChurchillAustralia

Personalised recommendations