A factorial HMM approach to robust isolated digit recognition in background music

Ameya Nitin Deoras; Mark Hasegawa-Johnson

A factorial HMM approach to robust isolated digit recognition in background music

Ameya Nitin Deoras, Mark Hasegawa-Johnson

Research output: Contribution to conference › Paper › peer-review

Abstract

This paper presents a novel solution to the problem of isolated digit recognition in background music. A Factorial Hidden Markov Model (FHMM) architecture is proposed to accurately model the simultaneous occurrence of two independent processes, such as an utterance of a digit and an excerpt of music. The FHMM is implemented with its equivalent HMM by extending Nadas' MIXMAX algorithm to a mixture of Gaussians PDF. At around 0 dB SNR, the proposed system shows an average relative reduction in word error rate of 57% in the recognition of isolated digits in background music.

Original language	English (US)
Pages	2093-2096
Number of pages	4
State	Published - 2004
Event	8th International Conference on Spoken Language Processing, ICSLP 2004 - Jeju, Jeju Island, Korea, Republic of Duration: Oct 4 2004 → Oct 8 2004

Other

Other	8th International Conference on Spoken Language Processing, ICSLP 2004
Country/Territory	Korea, Republic of
City	Jeju, Jeju Island
Period	10/4/04 → 10/8/04

ASJC Scopus subject areas

Language and Linguistics
Linguistics and Language

Library availability

Discover UIUC Full Text

Cite this

@conference{87a49a4ec05c490b96e00794a676de0d,

title = "A factorial HMM approach to robust isolated digit recognition in background music",

abstract = "This paper presents a novel solution to the problem of isolated digit recognition in background music. A Factorial Hidden Markov Model (FHMM) architecture is proposed to accurately model the simultaneous occurrence of two independent processes, such as an utterance of a digit and an excerpt of music. The FHMM is implemented with its equivalent HMM by extending Nadas' MIXMAX algorithm to a mixture of Gaussians PDF. At around 0 dB SNR, the proposed system shows an average relative reduction in word error rate of 57% in the recognition of isolated digits in background music.",

author = "Deoras, {Ameya Nitin} and Mark Hasegawa-Johnson",

note = "Funding Information: This work was supported in part by NSF award number 0132900. Statements in this paper reflect the opinions and conclusions of the authors, and are not endorsed by the NSF.; 8th International Conference on Spoken Language Processing, ICSLP 2004 ; Conference date: 04-10-2004 Through 08-10-2004",

year = "2004",

language = "English (US)",

pages = "2093--2096",

}

TY - CONF

T1 - A factorial HMM approach to robust isolated digit recognition in background music

AU - Deoras, Ameya Nitin

AU - Hasegawa-Johnson, Mark

N1 - Funding Information: This work was supported in part by NSF award number 0132900. Statements in this paper reflect the opinions and conclusions of the authors, and are not endorsed by the NSF.

PY - 2004

Y1 - 2004

N2 - This paper presents a novel solution to the problem of isolated digit recognition in background music. A Factorial Hidden Markov Model (FHMM) architecture is proposed to accurately model the simultaneous occurrence of two independent processes, such as an utterance of a digit and an excerpt of music. The FHMM is implemented with its equivalent HMM by extending Nadas' MIXMAX algorithm to a mixture of Gaussians PDF. At around 0 dB SNR, the proposed system shows an average relative reduction in word error rate of 57% in the recognition of isolated digits in background music.

AB - This paper presents a novel solution to the problem of isolated digit recognition in background music. A Factorial Hidden Markov Model (FHMM) architecture is proposed to accurately model the simultaneous occurrence of two independent processes, such as an utterance of a digit and an excerpt of music. The FHMM is implemented with its equivalent HMM by extending Nadas' MIXMAX algorithm to a mixture of Gaussians PDF. At around 0 dB SNR, the proposed system shows an average relative reduction in word error rate of 57% in the recognition of isolated digits in background music.

UR - http://www.scopus.com/inward/record.url?scp=85009074958&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85009074958&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85009074958

SP - 2093

EP - 2096

T2 - 8th International Conference on Spoken Language Processing, ICSLP 2004

Y2 - 4 October 2004 through 8 October 2004

ER -

A factorial HMM approach to robust isolated digit recognition in background music

Abstract

Other

ASJC Scopus subject areas

Library availability

Related links

Fingerprint

Cite this