DocumentCode
2037987
Title
Speaker identification by characteristics of linear prediction residual
Author
Fedorov, A.V. ; Omelchenko, Anatoliy V.
Author_Institution
Kharkov Nat. Univ. of Radio Electron., Kharkov, Ukraine
fYear
2013
fDate
8-14 Sept. 2013
Firstpage
450
Lastpage
451
Abstract
Synthesis and research of text-independent adaptive algorithms to identify speakers by characteristics of linear prediction residual are described. As the informative features we use coefficients of Fourier transform of an autocorrelation function of a linear prediction residual signal processed with different time window functions. Experimental tests on the proposed identification algorithms are carried out.
Keywords
Fourier transforms; adaptive signal detection; correlation theory; prediction theory; speaker recognition; Fourier transform; autocorrelation function; informative feature; linear prediction residual signal processing; speaker identification; text independent adaptive algorithm; time window function; Abstracts; Acoustic distortion; Adaptive algorithms; Educational institutions; Electronic mail; Fourier transforms; Speech;
fLanguage
English
Publisher
ieee
Conference_Titel
Microwave and Telecommunication Technology (CriMiCo), 2013 23rd International Crimean Conference
Conference_Location
Sevastopol
Print_ISBN
978-966-335-395-1
Type
conf
Filename
6652900
Link To Document