|
|
Multi-class classifier of non-speech audio based
on Fisher kernel |
Rongyan WANG,Gang LIU,Jun GUO,Yu FANG, |
Pattern Recognition
and Intelligent System Laboratory, Beijing University of Posts and
Telecommunications, Beijing 100876, China; |
|
|
Abstract Traditional multi-class classification methods based on Fisher kernel combine generative models such as Gaussian mixture models (GMMs) of all the classes together. However, the combination generates high dimensional feature vectors and leads to large computation. In this paper, a new classification method is proposed. This method adopts an intelligent feature space selection strategy by clustering similar Gaussian mixtures in order to reduce the feature dimensions. Audio classification experiments show that the proposed method is more accurate and effective with less computation compared with traditional methods.
|
Keywords
Fisher kernel
support vector machine (SVM)
Gaussian mixture model (GMM)
mixture clustering
|
Issue Date: 05 March 2010
|
|
|
Wold E, Blum T, Keislar D, Wheaton J. Content-basedclassification, search and retrieval of audio. IEEE MultiMedia, 1996, 3(3): 27―36
doi: 10.1109/93.556537
|
|
Rice S V. Audio and video retrieval based on audio content. Comparisonics. White Paper, 1998
|
|
Shirazi J, Ghaemmaghami S, Razzazi F. Improvements in audio classificationbased on sinusoidal modeling. In: Proceedingsof 2008 IEEE International Conference on Multimedia and Expo. 2008, 1485―1488
|
|
Pan W J, Yao Y, Liu Z J, Huang W Y. Audio classification in a weighted SVM. In: Proceedings of International Symposium on Communications andInformation Technologies. 2007, 468―472
|
|
Li X L, Du Z D, Zhang Y F. Kernel-based audio classification. In: Proceedings of 2006 International Conferenceon Machine Learning and Cybernetics. 2006, 3313―3316
|
|
Slaney M. Mixtures of probability experts for audio retrieval andindexing. In: Proceedings of 2002 IEEEInternational Conference on Multimedia and Expo. 2002, 1: 345―348
|
|
Guo G D, Li S Z. Content-basedaudio classification and retrieval by support vector machines. IEEE Transactions on Neural Networks, 2003, 14(1): 209―215
doi: 10.1109/TNN.2002.806626
|
|
Giannakopoulos T, Pikrakis A, Theodoridis S. A multi-class audio classificationmethod with respect to violent content in movies using Bayesian networks. In: Proceedings of IEEE the 9th Workshop on MultimediaSignal Processing. 2007, 90―93
|
|
Rabaoui A, Kadri H, Lachiri Z, Ellouze N. Usingrobust features with multi-class SVMs to classify noisy sounds. In: Proceedings of the 3rd International Symposiumon Communications, Control and Signal Processing. 2008, 594―599
|
|
Jaakkola T, Diekhans M, Haussler D. A discriminative frameworkfor detecting remote protein homologies. Journal of Computational Biology, 2000, 7(1―2): 95―114
doi: 10.1089/10665270050081405
|
|
Jaakkola T S, Haussler D. Exploitinggenerative models in discriminative classifiers. In: Solla S A, Leen T K, Müller K R, eds. Advances in Neural Information Processing Systems. Cambridge: MIT Press, 1999, 487―493
|
|
Smith N D, Gales M J F. Using SVMsto Classify Variable Length Speech Patterns. Technical Report CUED/F-INFENG/TR.412. 2001
|
|
Fine S, Navrátil J, Gopinath R A. A hybrid GMM/SVM approachto speaker identification. In: Proceedingsof 2001 IEEE International Conference on Acoustics, Speech, and SignalProcessing. 2001, 1: 417―420
|
|
Chen L, Man H, Nefian A V. Face recognition based on multi-classmapping of Fisher scores. Pattern Recognition, 2005, 38(6): 799―811
doi: 10.1016/j.patcog.2004.11.003
|
|
Aran O, Akarun L. Multi-classclassification strategies for Fisher scores of gesture and sign sequences. In: Proceedings of the 19th International Conferenceon Pattern Recognition. 2008, 1―4
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|