Presentation 2008-06-12
A Study of Speech Features Extraction Used for Robust Speech Recognition Using RSF/DRA
Mon SATO, Shingo YOSHIZAWA, Yoshikazu MIYANAGA, Noboru HAYASAKA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In many speech recognition systems, the speech features are based on MFCC. Exactly speaking, the 1st-12th MFCC, the 0th-12th ΔMFCC and the 0th-12th ΔΔMFCC are used into these systems. Under noisy condition, the spectral subtraction (SS) and the cepstrum mean subtraction (CMS) are widely used and the 0th MFCC is not used into these methods since the recognition accuracy is declined. As the new noise robust speech feature extraction, i.e., the running spectrum filtering (RSF) and the dynamic range adjustment (DRA) have been developed. In this report, the role of the 0th MFCC is investigated into RSF and DRA. The evaluation on this speech feature is shown in some experiments.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) RSF / DRA / Robust / Speech Recognition
Paper # SIS2008-8
Date of Issue

Conference Information
Committee SIS
Conference Date 2008/6/5(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 Smart Info-Media Systems (SIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Study of Speech Features Extraction Used for Robust Speech Recognition Using RSF/DRA
Sub Title (in English)
Keyword(1) RSF
Keyword(2) DRA
Keyword(3) Robust
Keyword(4) Speech Recognition
1st Author's Name Mon SATO
1st Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University()
2nd Author's Name Shingo YOSHIZAWA
2nd Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
3rd Author's Name Yoshikazu MIYANAGA
3rd Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
4th Author's Name Noboru HAYASAKA
4th Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
Date 2008-06-12
Paper # SIS2008-8
Volume (vol) vol.108
Number (no) 85
Page pp.pp.-
#Pages 6
Date of Issue