Presentation 2005/4/14
Local Vector Coding for Voiced Speech Signals
Masashi ITO, Masafumi YANO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Line-type spectra are observed in frequency responses for voiced sound and sound of some musical instrument. It is difficult to analyze these types of sounds accurately by using power spectrogram because these sounds are usually unstationary. We have proposed a new method, termed local vector coding (LVC), to analyze these sounds. LVC can estimate not only instantaneous amplitude and frequency but also their local velocity for each component. In this article, we examine LVC for naturally uttered speech input. The mean error between the inputs and the estimated spectra is 23.7dB in S/N. From the obtained results, the problems and validities of the LVC method are discussed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Local vector coding / voice / acoustical analysis / quasi-periodic signal
Paper # EA2005-2,SP2005-2
Date of Issue

Conference Information
Committee EA
Conference Date 2005/4/14(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 Engineering Acoustics (EA)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Local Vector Coding for Voiced Speech Signals
Sub Title (in English)
Keyword(1) Local vector coding
Keyword(2) voice
Keyword(3) acoustical analysis
Keyword(4) quasi-periodic signal
1st Author's Name Masashi ITO
1st Author's Affiliation Research Institute of Electrical Communication, Tohoku University()
2nd Author's Name Masafumi YANO
2nd Author's Affiliation Research Institute of Electrical Communication, Tohoku University
Date 2005/4/14
Paper # EA2005-2,SP2005-2
Volume (vol) vol.105
Number (no) 9
Page pp.pp.-
#Pages 6
Date of Issue