Presentation 2008-06-26
Direct LSP Estimation Using Recursive Gauss-Newton Method with a Tapered Window
Gyo SATO, Takefumi KITAYAMA, Atsushi NAKAGAKI, Koji SHIBATA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The LSP is one of the most important feature parameters of speech based on the LPC. An adaptive LSP estimation method using the recursive Gauss-Newton method has been proposed. The exponential or rectangular window is usually introduced into its least squares criterion to estimate time-dependent LSP parameters. In this paper, two types of tapered windows, that is, Hanning and modified Barnwell windows, are applied to the adaptive LSP estimation to improve the steady-state behavior as well as the tracking behavior.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) LSP / adaptive estimation / resursive Gauss-Newton method / tapered widnow
Paper # CAS2008-3,VLD2008-16,SIP2008-37
Date of Issue

Conference Information
Committee CAS
Conference Date 2008/6/19(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Circuits and Systems (CAS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Direct LSP Estimation Using Recursive Gauss-Newton Method with a Tapered Window
Sub Title (in English)
Keyword(1) LSP
Keyword(2) adaptive estimation
Keyword(3) resursive Gauss-Newton method
Keyword(4) tapered widnow
1st Author's Name Gyo SATO
1st Author's Affiliation Faculty of Engineering, Kitami Institute of Technology()
2nd Author's Name Takefumi KITAYAMA
2nd Author's Affiliation Faculty of Engineering, Kitami Institute of Technology
3rd Author's Name Atsushi NAKAGAKI
3rd Author's Affiliation Faculty of Engineering, Kitami Institute of Technology
4th Author's Name Koji SHIBATA
4th Author's Affiliation Faculty of Engineering, Kitami Institute of Technology
Date 2008-06-26
Paper # CAS2008-3,VLD2008-16,SIP2008-37
Volume (vol) vol.108
Number (no) 104
Page pp.pp.-
#Pages 4
Date of Issue