Abstract
A modified GMM with an embedded TDNN is proposed to speaker recognition. The model integrates the merits of GMM and TDNN. TDNN is used to digest the time information of the feature sequences, and through the transformation of the feature vectors the model makes the hypothesis of variable independence which maximum likelihood needed more reasonable. In the process of training, GMM and TDNN are trained as a whole and the parameters of GMM and TDNN are updated alternately. Experiments show that the proposed model improves accuracy rate against baseline GMM at all SNR with a maximum to 22%.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Bimbot, F., Bonastre, J.F., et al.: A Tutorial on Text-Independent Speaker Verification. EURASIP Journal on Applied Signal Processing 4, 430–451 (2004)
Reynolds, D.A., Rose, R.C.: Robust Text-independent Speaker Identification Using Gaussian Mixture Speaker Models. IEEE Trans. Speech Audio Processing 3, 72–83 (1995)
Reynolds, D., Quatieri, T., Dunn, R.: Speaker Verification Using Adapted Gaussian Mixture Models. Digital Signal Processing 10, 19–41 (2000)
Campbell, S.D.E., Reynolds, D.A.: SVM based speaker verification using A GMM supervector kernel and NAP variability compensation. In: Proceedings of ICASSP, pp. 97–100 (2006)
Yin, S.-C., Rose, R., Kenny, P.: A joint factor analysis approach to progressive model adaptation in Text-Independent speaker verification. IEEE Trans. on Audio, Speech and Language Processing 15, 1999–2110 (2007)
Dehak, N., Dehak, R., Kenny, P., Dumouchel, P.: Comparison between Factor Analysis and GMM Support Vector Machines for Speaker Verification. In: Proceedings of IEEE Odyssey 2008 - The Speaker and Language Recognition Workshop (Odyssey 2008), Stellenbosch, South Africa, January 21-25 (2008)
Mak, M.W., Allen, W.G., Sexton, G.G.: Speaker identification using multilayer perceptron and radial basis function networks. Neurocomputing 6(1), 99–117 (1994)
Murty, K.S.R., Yegnanarayana, B.: Combining evidence from residual phase and MFCC features for speaker recognition. IEEE signal processing letters 13, 52–55 (2006)
Bennani, Y., Gallinari, P.: On the use of TDNN-Extracted Features Information In Talker Identification. In: Proceedings of ICASSP, pp. 385–388 (1991)
Dev, A.: Effect of retroflex sounds on the recognition of Hindi voiced and unvoiced stops. AI & SOCIETY 23(4), 603–612 (2009)
Zhou, H., Taj, M., Cavallaro, A.: Target Detection and Tracking With Heterogeneous Sensors. IEEE Journal of Selected Topics in Signal Processing 2, 503–513 (2008)
Vogl, T.P., Mangis, J.K., Aigler, A.K.: Accelerating the convergence of the backprogagating method. Biological Cybernetics 59, 256–264 (1988)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Chen, C., Zhao, L. (2009). Speaker Recognition Based on GMM with an Embedded TDNN. In: Leung, C.S., Lee, M., Chan, J.H. (eds) Neural Information Processing. ICONIP 2009. Lecture Notes in Computer Science, vol 5864. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-10684-2_83
Download citation
DOI: https://doi.org/10.1007/978-3-642-10684-2_83
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-10682-8
Online ISBN: 978-3-642-10684-2
eBook Packages: Computer ScienceComputer Science (R0)