HMM adaptation using statistical linear approximation for robust automatic speech recognition

Michael Berkovitch; Ilan D. Shallom

HMM adaptation using statistical linear approximation for robust automatic speech recognition

Michael Berkovitch, Ilan D. Shallom

Department of Electrical & Computer Engineering

Research output: Contribution to journal › Conference article › peer-review

1 Scopus citations

Abstract

The lack of noise robustness is one of the main drawbacks of an Automatic Speech Recognition (ASR) system. A well trained ASR system can achieve high recognition rate on quiet laboratory conditions, but perform poorly in real life environments. In this paper we will present a noise robustness method which uses the clean speech Hidden Markov Models (HMM) and noise statistics, to create an approximation of the degraded speech HMM using the Statistical Linear Approximation (SLA). Experiments using the proposed methods had shown up to 87.7% word error rate improvement.

Original language	English
Pages (from-to)	1301-1304
Number of pages	4
Journal	Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
State	Published - 1 Dec 2008
Event	INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association - Brisbane, QLD, Australia Duration: 22 Sep 2008 → 26 Sep 2008

Keywords

Robust ASR
Speech recognition

ASJC Scopus subject areas

Human-Computer Interaction
Signal Processing
Software
Sensory Systems

Cite this

@article{1a873ad1b4c546f285a0008c6452dddc,

title = "HMM adaptation using statistical linear approximation for robust automatic speech recognition",

abstract = "The lack of noise robustness is one of the main drawbacks of an Automatic Speech Recognition (ASR) system. A well trained ASR system can achieve high recognition rate on quiet laboratory conditions, but perform poorly in real life environments. In this paper we will present a noise robustness method which uses the clean speech Hidden Markov Models (HMM) and noise statistics, to create an approximation of the degraded speech HMM using the Statistical Linear Approximation (SLA). Experiments using the proposed methods had shown up to 87.7% word error rate improvement.",

keywords = "Robust ASR, Speech recognition",

author = "Michael Berkovitch and Shallom, {Ilan D.}",

year = "2008",

month = dec,

day = "1",

language = "English",

pages = "1301--1304",

journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",

issn = "2308-457X",

note = "INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association ; Conference date: 22-09-2008 Through 26-09-2008",

}

TY - JOUR

T1 - HMM adaptation using statistical linear approximation for robust automatic speech recognition

AU - Berkovitch, Michael

AU - Shallom, Ilan D.

PY - 2008/12/1

Y1 - 2008/12/1

N2 - The lack of noise robustness is one of the main drawbacks of an Automatic Speech Recognition (ASR) system. A well trained ASR system can achieve high recognition rate on quiet laboratory conditions, but perform poorly in real life environments. In this paper we will present a noise robustness method which uses the clean speech Hidden Markov Models (HMM) and noise statistics, to create an approximation of the degraded speech HMM using the Statistical Linear Approximation (SLA). Experiments using the proposed methods had shown up to 87.7% word error rate improvement.

AB - The lack of noise robustness is one of the main drawbacks of an Automatic Speech Recognition (ASR) system. A well trained ASR system can achieve high recognition rate on quiet laboratory conditions, but perform poorly in real life environments. In this paper we will present a noise robustness method which uses the clean speech Hidden Markov Models (HMM) and noise statistics, to create an approximation of the degraded speech HMM using the Statistical Linear Approximation (SLA). Experiments using the proposed methods had shown up to 87.7% word error rate improvement.

KW - Robust ASR

KW - Speech recognition

UR - http://www.scopus.com/inward/record.url?scp=84867193149&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:84867193149

SN - 2308-457X

SP - 1301

EP - 1304

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

T2 - INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association

Y2 - 22 September 2008 through 26 September 2008

ER -

HMM adaptation using statistical linear approximation for robust automatic speech recognition

Abstract

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this