Speaker Characterization Using Principal Component Analysis and Wavelet Transform for Speaker Verification

摘要

In this paper, we investigate the use of the Wavelet Transform for text-dependent and text-independent Speaker Verification tasks. We have introduced a Principal Component Analysis based wavelet transform to perform frequencies segmentation with levels decomposition. A speaker dependent library tree has been built, corresponding to the best structure for a given speaker. The constructed tree is abstract and specific to every single speaker. Therefore the extracted parameters are more discriminative and appropriate for speaker verification applications. It has been compared to MFCC's and other wavelet-based parameters. Experiments have been conducted using corpus, extracted from Yoho and Spidre Databases. This technique has shown robustness and 100% efficiency in both cases.