Partition Based Feature Processing for Improved Music Classification

  • Igor Vatolkin
  • Wolfgang Theimer
  • Martin Botteck
Conference paper
Part of the Studies in Classification, Data Analysis, and Knowledge Organization book series (STUDIES CLASS)


Identifying desired music amongst the vast amount of tracks in today’s music collections has become a task of increasing attention for consumers. Music classification based on perceptual features promises to help sorting a collection according to personal music categories determined by the user’s personal taste and listening habits. Regarding limits of processing power and storage space available in real (e.g. mobile) devices necessitates to reduce the amount of feature data used by such classification. This paper compares several methods for feature pruning– experiments on realistic track collections show that an approach attempting to identify relevant song partitions not only allows to reduce the amount of processed feature data by 90% but also helps to improve classification accuracy. They indicate that a combination of structural information and temporal continuity processing of partition based classification helps to substantially improve overall performance.


Classification Performance Gaussian Mixture Model Audio Feature Personal Taste Accuracy Rank 
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.



We thank the Klaus Tschira Foundation for financial support.


  1. AllMusicGuide (2010) Webpage. URL, cited Feb 2011
  2. Blume H, Haller M, Botteck M, Theimer W (2008) Perceptual feature based music classification - a DSP perspective for a new type of application. In: Proceedings of the SAMOS VIII Conference, (IC-SAMOS), Samos, Greece, pp 92–99Google Scholar
  3. Klapuri AP, Eronen AJ, Astola JT (2006) Analysis of the meter of acoustic musical signals. IEEE Trans Audio Speech Audio Process 14(1):342–355CrossRefGoogle Scholar
  4. Lartillot O, Toiviainen P, Eerola T (2008) A Matlab toolbox for music information retrieval. In: Preisach C, Burkhardt H, Schmidt-Thieme L, Decker R (eds) Data analysis, machine learning and knowledge organization. Springer, BerlinGoogle Scholar
  5. Mörchen F, Ultsch A, Nöcker M, Stamm C (2005) Databionic visualization of music collections according to perceptual distance. In: Proceedings of the 6th International Conference on Music Information Retrieval, London, UK, pp 396–403Google Scholar
  6. Music Test Database of the CI Research Group (2010) Cited 30 Jul 2010
  7. Pachet F, Cazaly D (2000) A taxonomy of musical genres. In: Proc. Content-Based Multimedia Information Access, (RIAO), Paris, FranceGoogle Scholar
  8. Paulus J, Klapuri A (2008) Music structure analysis using a probabilistic fitness measure and an integrated musicological model. In: Proc. of the 9th Int’l Conf. on Music Information RetrievalGoogle Scholar
  9. Theimer W, Vatolkin I, Botteck M, Buchmann M (2008a) Content-based similarity search and visualization for personal music categories. In: Proc. of the Sixth International Workshop on Content-Based Multimedia Indexing, LondonGoogle Scholar
  10. Theimer W, Vatolkin I, Eronen A (2008b) Definitions of Audio Features for Music Content Description. Technical Report TR08-2-001, Chair of Algorithm Engineering, TU DortmundGoogle Scholar
  11. Vatolkin I, Theimer W, Rudolph G (2009) Design and comparison of different evolution strategies for feature selection and consolidation in music classification. In: Proc. of the 2009 IEEE Congress on Evolutionary Computation (CEC), IEEE Press, Piscataway, NJ, pp 174–181Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Igor Vatolkin
    • 1
  • Wolfgang Theimer
    • 2
  • Martin Botteck
  1. 1.TU DortmundDortmundGermany
  2. 2.Research In MotionBochumGermany

Personalised recommendations