Wavelet feature selection based neural networks with application to the text independent speaker identification

A wavelet packets feature selection derived by using neuro-fuzzy evaluation index for speaker identification is described. The concept of a flexible membership function incorporating weighed distance is introduced in the evaluation index to make the modeling of clusters more appropriate. Experimental evaluation of the systems performance was conducted on three speech databases. Our results have shown that this feature selection introduced better performance than the wavelet features with respect to the percentages of recognition.

[1]  Sung-Yang Bang,et al.  Feature selection for multi-class classification using pairwise class discriminatory measure and covering concept , 2000 .

[2]  Shung-Yung Lung.  Feature extracted from wavelet eigenfunction estimation for text-independent speaker recognition , 2004, Pattern Recognit..

[3]  Alfredo Petrosino,et al.  Encoding nondeterministic fuzzy tree automata into recursive neural networks , 2004, IEEE Transactions on Neural Networks.

[4]  Harvey F. Silverman,et al.  Time-varying feature selection and classification of unvoiced stop consonants , 1994, IEEE Trans. Speech Audio Process..

[5]  J. N. Gowdy,et al.  Feature extraction using discrete wavelet transform for speech recognition , 2000, Proceedings of the IEEE SoutheastCon 2000. 'Preparing for The New Millennium' (Cat. No.00CH37105).

[6]  Shung-Yung Lung.  Wavelet Feature Selection Using Fuzzy Approach to Text Independent Speaker Recognition , 2005, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[7]  Nashville Airport Marriott.  PROCEEDINGS OF The IEEE Southeastcon 2000 , 2000 .

[8]  Mübeccel Demirekler,et al.  Speaker identification through use of features selected using genetic algorithm , 1998 .

[9]  Sungwook Chang,et al.  Speech feature extracted from adaptive wavelet for speech recognition , 1998 .

[10]  Chi-Chung Cheung,et al.  Magnified gradient function with deterministic weight modification in adaptive learning , 2004, IEEE Transactions on Neural Networks.

[11]  Shung Yung Lung.  APPLIED MULTI-WAVELET FEATURE TO TEXT INDEPENDENT SPEAKER IDENTIFICATION , 2004 .

[12]  Douglas A. Reynolds,et al.  Robust text-independent speaker identification using Gaussian mixture speaker models , 1995, IEEE Trans. Speech Audio Process..