Separation of excitation source and vocal tract transfer function via an MVDR analysis of speech

Author

Santarelli, Alberto ; Omologo, Maurizio ; Armani, Luca

Author_Institution

ITC-irst, Trento, Italy

fYear

2003

fDate

19-22 Oct. 2003

Firstpage

115

Lastpage

118

Abstract

The usage of MVDR (minimum variance distortionless response) spectrum estimation for separating excitation source from vocal tract transfer function in speech analysis is discussed. It is shown how an MVDR all-pole model and its inverse filter for source deconvolution can easily be derived by applying spectral factorization. We outline the theory of the Bauer method for spectral factorization, presenting examples of the application to real data recorded by distant microphones and giving insights on this technique.

Keywords

FIR filters; deconvolution; parameter estimation; poles and zeros; signal reconstruction; spectral analysis; speech processing; transfer functions; Bauer method; FIR filter; MVDR spectrum estimation; all-pole model; excitation source separation; inverse filter; source deconvolution; spectral factorization; speech analysis; speech reconstruction; vocal tract transfer function; Equations; Filters; Inverse problems; Microphones; Predictive models; Resonance; Speech analysis; Speech processing; Transfer functions; Transforms;

fLanguage

English

Publisher

ieee

Conference_Titel

Applications of Signal Processing to Audio and Acoustics, 2003 IEEE Workshop on.

Print_ISBN

0-7803-7850-4

Type

conf

DOI

10.1109/ASPAA.2003.1285833

Filename

1285833