Advertisement

Comparing Probabilistic Models for Melodic Sequences

  • Athina Spiliopoulou
  • Amos Storkey
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6913)

Abstract

Modelling the real world complexity of music is a challenge for machine learning. We address the task of modeling melodic sequences from the same music genre. We perform a comparative analysis of two probabilistic models; a Dirichlet Variable Length Markov Model (Dirichlet-VMM) and a Time Convolutional Restricted Boltzmann Machine (TC-RBM). We show that the TC-RBM learns descriptive music features, such as underlying chords and typical melody transitions and dynamics. We assess the models for future prediction and compare their performance to a VMM, which is the current state of the art in melody generation. We show that both models perform significantly better than the VMM, with the Dirichlet-VMM marginally outperforming the TC-RBM. Finally, we evaluate the short order statistics of the models, using the Kullback-Leibler divergence between test sequences and model samples, and show that our proposed methods match the statistics of the music genre significantly better than the VMM.

Keywords

melody modeling music feature extraction time convolutional restricted Boltzmann machine variable length Markov model Dirichlet prior 

References

  1. Ackley, D.H., Hinton, G.E., Sejnowski, T.J.: A learning algorithm for Boltzmann machines. Cognitive Science 9(1), 147–169 (1985)CrossRefGoogle Scholar
  2. Dubnov, S., Assayag, G., Lartillot, O., Bejerano, G.: Using machine-learning methods for musical style modeling. Computer 36(10), 73–80 (2003)CrossRefGoogle Scholar
  3. Eck, D., Lapalme, J.: Learning musical structure directly from sequences of music. Technical report, Université de Montreal (2008)Google Scholar
  4. Eck, D., Schmidhuber, J.: Learning the long-term structure of the blues. In: Dorronsoro, J.R. (ed.) ICANN 2002. LNCS, vol. 2415, pp. 284–289. Springer, Heidelberg (2002)CrossRefGoogle Scholar
  5. Eerola, T., Toiviainen, P.: MIDI Toolbox: MATLAB Tools for Music Research. University of Jyväskylä, Jyväskylä, Finland (2004), www.jyu.fi/musica/miditoolbox/
  6. Hinton, G.E.: Training products of experts by minimizing contrastive divergence. Neural Computation 14(8), 1771–1800 (2002)CrossRefzbMATHGoogle Scholar
  7. Hinton, G.E., Osindero, S., Teh, Y.W.: A fast learning algorithm for deep belief nets. Neural Computation 18(7), 1527–1554 (2006)CrossRefzbMATHMathSciNetGoogle Scholar
  8. Lavrenko, V., Pickens, J.: Polyphonic music modeling with random fields. In: Rowe, L.A., Vin, H.M., Plagemann, T., Shenoy, P.J., Smith, J.R. (eds.) Proceedings of the Eleventh ACM International Conference on Multimedia, ACM Multimedia, pp. 120–129. ACM, New York (2003)CrossRefGoogle Scholar
  9. Lee, H., Ekanadham, C., Ng, A.Y.: Sparse deep belief net model for visual area V2. In: Platt, J.C., Koller, D., Singer, Y., Roweis, S.T. (eds.) NIPS. Advances in NIPS, vol. 20. MIT Press, Cambridge (2008)Google Scholar
  10. Lee, H., Grosse, R., Ranganath, R., Ng, A.Y.: Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations. In: Danyluk, A.P., Bottou, L., Littman, M.L. (eds.) ICML. ACM ICPS, vol. 382, p. 77. ACM, New York (2009)Google Scholar
  11. Lerdahl, F., Jackendoff, R.: A Generative Theory of Tonal Music. The MIT Press, Cambridge (1983)Google Scholar
  12. Norouzi, M., Ranjbar, M., Mori, G.: Stacks of convolutional restricted Boltzmann machines for shift-invariant feature learning. In: IEEE Computer Society Conference on CVPR, CVRP 2009, pp. 2735–2742. IEEE, Los Alamitos (2009)Google Scholar
  13. Paiement, J.-F.: Probabilistic Models for Music. PhD thesis, Ecole Polytechnique Fédérale de Lausanne (EPFL) (2008)Google Scholar
  14. Ron, D., Singer, Y., Tishby, N.: The power of amnesia. In: Cowan, J.D., Tesauro, G., Alspector, J. (eds.) NIPS. Advances in NIPS, vol. 6, pp. 176–183. Morgan Kaufmann, San Francisco (1994)Google Scholar
  15. Sutskever, I., Hinton, G.E.: Learning multilevel distributed representations for high-dimensional sequences. Journal of ML Research - Proceedings Track 2, 548–555 (2007)Google Scholar
  16. Taylor, G.W., Hinton, G.E.: Factored conditional restricted Boltzmann machines for modeling motion style. In: Danyluk, A.P., Bottou, L., Littman, M.L. (eds.) ICML. ACM ICPS, vol. 382, p. 129. ACM, New York (2009)Google Scholar
  17. Taylor, G.W., Hinton, G.E., Roweis, S.T.: Modeling human motion using binary latent variables. In: Schölkopf, B., Platt, J.C., Hoffman, T. (eds.) NIPS. Advances in NIPS, vol. 19, pp. 1345–1352. MIT Press, Cambridge (2007)Google Scholar
  18. Weiland, M., Smaill, A., Nelson, P.: Learning musical pitch structures with hierarchical hidden Markov models. Technical report, University of Edinburgh (2005)Google Scholar
  19. Welling, M., Rosen-Zvi, M., Hinton, G.E.: Exponential family harmoniums with an application to information retrieval. In: NIPS. Advances in NIPS, vol. 17 (2004)Google Scholar
  20. Wood, F., Archambeau, C., Gasthaus, J., James, L., Teh, Y.W.: A stochastic memoizer for sequence data. In: Danyluk, A.P., Bottou, L., Littman, M.L. (eds.) ICML. ACM ICPS, vol. 382, p. 142. ACM, New York (2009)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Athina Spiliopoulou
    • 1
  • Amos Storkey
    • 1
  1. 1.School of InformaticsUniversity of EdinburghUnited Kingdom

Personalised recommendations