Presentation 2005/12/15
A study on noisy-speech recognition performance using quantized LSP parameters based on AURORA-2J database
Yoshinori MORITA, Kousuke KAMEDA, Tetsuo FUNADA, Hideyuki NOMURA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In distributed speech recognition (DSR), it is recommended that the MFCC (Mel Frequency Cepstrum Coefficient) is used for speech analysis. However, LSP is widely used for quantity of characteristic in speech encoding. Therefore, it is necessary to examine performance of speech recognition when LSP is used. As a result of having compared LSP with MFCC by Aurora-2J database, a recognition rate of 80.0% are provided with multicondition, relatively becomes -42.2%. And a recognition rate of 70.5% is provided in 12 bits quantization.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) LSP / AURORA2J / Vector Quantization
Paper # NLC2005-86,SP2005-119
Date of Issue

Conference Information
Committee NLC
Conference Date 2005/12/15(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 Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A study on noisy-speech recognition performance using quantized LSP parameters based on AURORA-2J database
Sub Title (in English)
Keyword(1) LSP
Keyword(2) AURORA2J
Keyword(3) Vector Quantization
1st Author's Name Yoshinori MORITA
1st Author's Affiliation Ishikawa N.C.T.()
2nd Author's Name Kousuke KAMEDA
2nd Author's Affiliation Faculty of Engineering, Kanazawa University
3rd Author's Name Tetsuo FUNADA
3rd Author's Affiliation Faculty of Engineering, Kanazawa University
4th Author's Name Hideyuki NOMURA
4th Author's Affiliation Faculty of Engineering, Kanazawa University
Date 2005/12/15
Paper # NLC2005-86,SP2005-119
Volume (vol) vol.105
Number (no) 494
Page pp.pp.-
#Pages 2
Date of Issue