NON DIRECTLY ACOUSTIC PROCESS FOR COSTLESS SPEAKER RECOGNITION AND INDEXATION
作者： Corinne Fredouille , Jean-François Bonastre , Teva Merlin
DOI:
关键词:
摘要: This paper presents a new approach to speaker recognition and indexation systems, based on non-directly-acoustic processing. method is specifically designed lower the complexity of modeling phase, compared classical techniques, as well decrease required amount learning data, making it particularly well-suited on-line (needed for indexation) use embedded systems.
archives-ouvertes.fr
researchgate.net 参考文章(5)
Guillaume Gravier, Chafic Mokbel, Gérard Chollet, Model dependent spectral representations for speaker recognition. conference of the international speech communication association. ,(1997)
D. Charlet, D. Jouvet, Optimizing feature set for speaker verification Pattern Recognition Letters. ,vol. 18, pp. 873- 879 ,(1997) , 10.1016/S0167-8655(97)00064-0
Frédéric Bimbot, Ivan Magrin-Chagnolleau, Luc Mathan, Second-order statistical measures for text-independent speaker identification Speech Communication. ,vol. 17, pp. 177- 192 ,(1995) , 10.1016/0167-6393(95)00013-E
L.F. Lamel, J.L. Gauvain, Speaker verification over the telephone Speech Communication. ,vol. 31, pp. 141- 154 ,(2000) , 10.1016/S0167-6393(99)00075-8
D.A. Reynolds, R.C. Rose, Robust text-independent speaker identification using Gaussian mixture speaker models IEEE Transactions on Speech and Audio Processing. ,vol. 3, pp. 72- 83 ,(1995) , 10.1109/89.365379