Research of Verification-Mode Voice Recognition Based on UBM-GMM
Abstract:
This study proposes a Verification-Mode voice recognition system which uses in IC card. Improvements have been made both in the two periods of voice recognition: on the one hand, a new method of endpoint detection is used in voice pretreatment and doing cepstrum filtering for coefficients in the MFCC extraction; on the other hand, use UBM-GMM algorithm in the process of template matching and make it better to the verification mode. Experiment with putting the users voice data into IC card to do the user verification and the result shows that this mode eliminates the dependence of voice text and improve the speed and precision of the identification.
How to cite this article
Fang Zhi-Gang, Sun Chao and Xu Jie, 2013. Research of Verification-Mode Voice Recognition Based on UBM-GMM. Information Technology Journal, 12: 5661-5664.
REFERENCES
Canny, J., 1986. A computational approach to edge detection. IEEE Trans. Pattern Anal. Mach. Intell., 8: 679-698.
CrossRef Direct Link
Reynolds, D.A. and R.С. Rose, 1995. Robust text-independent speaker identification using gaussian mixture speaker models. IEEE Trans. Speech Audio Process., 3: 72-83.
Direct Link
Higgins, A.L., L. Bahler and J. Porter, 1991. Speaker verification using randomized phrase prompting. Dig. Sig. Process., 1: 89-106.
CrossRef Direct Link
Sambur, M., 1975. Selection of acoustic features for speaker identification. IEEE Trans. Acoustics Speech Signal Process., 23: 176-182.
CrossRef
Zhang, Z., C. Chen, J. Sun and K.L. Chan, 2003. EM Algorithms for Gaussian mixtures with split-and-merge operation. Patt. Recog., 36: 1973-1983.
CrossRef
© Science Alert. All Rights Reserved