Music retrieval based on a multi-samples selection strategy for support vector machine active learning

ABSTRACT

In active learning based music retrieval systems, providing multiple samples to the user for feedback is very necessary. In this paper, we present a new multi-samples selection strategy designed for support vector machine active learning. Aiming to reduce the redundancy between the selected samples, the strategy enforces the selected samples to be diverse by explicitly maximizing the distance between each other in the feature space. Experimental results on a music genre database demonstrated the effectiveness of the proposed strategy in selecting relevant multiple samples for human feedback on them.

REFERENCES

• 1: K. Brinker. Incorporating diversity in active learning with support vector machines. In Proceedings of the 20th international conference on machine learning, pages 59--66, 2003.
• 2: C. Chang and C. Lin. LIBSVM: a library for support vector machines. http://www.csie.ntu.edu.tw/~cjlin/libsvm, 2001.
• 3: M. Ferecatu, M. Crucianu, and N. Boujemaa. Sample selection strategies for relevance feedback in region-based image retrieval. In Proceedings of pacific-rim conference on multimedia, 2004.
• 4: P. Herrera, J. Bello, and G. W. et al. Simac: semantic interaction with music audio contents. In Proceedings of the 2nd european workshop on the integration of knowledge, semantics and digital media technology, pages 399--406, 2005.
• 5: M. Mandel, G. Poliner, and D. Ellis. Support vector machine active learning for music retrieval. Multimedia systems, 12(1):3--13, 2006.
• 6: Nicholas Roy , Andrew McCallum, Toward Optimal Active Learning through Sampling Estimation of Error Reduction, Proceedings of the Eighteenth International Conference on Machine Learning, p.441-448, June 28-July 01, 2001
• 7: T. Mitchell. Generalization as search. Artificial Intelligence, pages 203--226, 1982.
• 8: Simon Tong , Daphne Koller, Support vector machine active learning with applications to text classification, The Journal of Machine Learning Research, 2, p.45-66, 3/1/2002 [doi>10.1162/153244302760185243]

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.5 INFORMATION INTERFACES AND PRESENTATION (I.7)
      H.5.5 Sound and Music Computing

General Terms:
Algorithms, Design, Experimentation

Keywords:
active learning, music retrieval, relevance feedback, selection strategy, support vector machine