A Novel Approach for Speaker Recognition by Using Wavelet Analysis and Support Vector Machines
Speaker recognition approach through wavelet analysis as well as support vector machines is presented in this paper. The wavelet-based approach is used to differentiate among regular and irregular voices. The wavelet filter banks were utilized to coincide by means of support vector machine for extraction of the feature and its classification. This approach creates utilization of wavelets as well as support vector machine to separate particular speech signal through multi-dialog settings. In this approach, first we apply the wavelets to calculate audio features those have sub-band power and calculated pitch values from the given data of the speech. Multi-speaker separation of speech data is carried out by the utilization of SVM more than these audio features as well as other values of the signal. This entire database was utilized to calculate the performance of the system and it represents over 95 % accuracy.
KeywordsSpeaker recognition Wavelet analysis Support vector machines
- 7.Wu, J.-D., Lin, B.-F.: Speaker identification using discrete wavelet packet transform technique with irregular decomposition. Expert Syst. Appl. 36(2), 3136–3143 (2009)Google Scholar
- 8.Francesco, P., Fiore, U., Alfredo, D.S.: On the detection of card-sharing traffic through wavelet analysis and Support Vector Machines. Appl. Soft Comput. 13(1), 615–627 (2013)Google Scholar
- 11.You, C.H., Li, H.: Relevance factor of maximum a posteriori adaptation for GMM–NAP–SVM in speaker and language recognition. Comput. Speech Lang. 30, 116–134 (2014)Google Scholar