Presentation	1995/9/12 Normalizing Lombard speech under different noise conditions Atsushi Wakao, Kazuya Takeda, Fumitada Itakura,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	In this paper, normalization of Lombard speech under different noise conditions is discussed. First, we show how Lombard speech is affected by difference of noise by conducting recognition experiments. Second, optimal normalization by regressive model is calculated for different speaker and noise conditions. The resultant spectral distance after the normalization show that the normalization by regressive model can reduce the distance more than 10% in average, and that the unvoiced stops are mostly affected by the difference of noise. Finally, by recognition experiments, we confirm that the normalization method can improve the recognition accuracy by up to 20%.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	Lombard effect / adaptation / DTW recognizer
Paper #
Date of Issue

Conference Information
Committee	DSP
Conference Date	1995/9/12(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	Digital Signal Processing (DSP)
Language	JPN
Title (in Japanese)	(See Japanese page)
Sub Title (in Japanese)	(See Japanese page)
Title (in English)	Normalizing Lombard speech under different noise conditions
Sub Title (in English)
Keyword(1)	Lombard effect
Keyword(2)	adaptation
Keyword(3)	DTW recognizer
1st Author's Name	Atsushi Wakao
1st Author's Affiliation	Graduate School of Engineering, Nagoya University()
2nd Author's Name	Kazuya Takeda
2nd Author's Affiliation	Graduate School of Engineering, Nagoya University
3rd Author's Name	Fumitada Itakura
3rd Author's Affiliation	Graduate School of Engineering, Nagoya University
Date	1995/9/12
Paper #
Volume (vol)	vol.95
Number (no)	227
Page	pp.pp.-
#Pages	8
Date of Issue