Presentation 2001/7/19
Quality Improvement of Speech by Neural Network Vector Quantization Combined with Spectral Subtraction Method
Yoshinori Morita, Tetsuo Funada,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In the mobile communication, the speech encoding performance in the environment for instance, with the background noise in the car etc.is paid to attention. In this paper the speech is encoded by the vector quantization by the use of the neural network. First of all, this method reduces the noise by the spectral subtraction method. At this time, ″musical noise″is generated and suppressed by using the identity mapping ability of the neural network and the expressed sound source. It was confirmed that the noise was suppressed to the speech of SNR 0dB by a preliminary listening experiment.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) speech coding / neural network / vector quantization / LSP / noise suppression
Paper # SP2001-38
Date of Issue

Conference Information
Committee SP
Conference Date 2001/7/19(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 Speech (SP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Quality Improvement of Speech by Neural Network Vector Quantization Combined with Spectral Subtraction Method
Sub Title (in English)
Keyword(1) speech coding
Keyword(2) neural network
Keyword(3) vector quantization
Keyword(4) LSP
Keyword(5) noise suppression
1st Author's Name Yoshinori Morita
1st Author's Affiliation Ishikawa National College of Technology()
2nd Author's Name Tetsuo Funada
2nd Author's Affiliation Faculty of Engineering, Kanazawa University
Date 2001/7/19
Paper # SP2001-38
Volume (vol) vol.101
Number (no) 232
Page pp.pp.-
#Pages 6
Date of Issue