Perceptual speech modeling for noisy speech recognition

Chung Hsien Wu; Yu Hsien Chiu; Huigan Lim

Perceptual speech modeling for noisy speech recognition

Chung Hsien Wu, Yu Hsien Chiu, Huigan Lim

Department of Computer Science and Information Engineering

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

Abstract

This paper proposes a perceptual modeling approach with a two-stage recognition to deal with the issues of recognition degradation in noisy environment. The auditory masking effect is used for speech enhancement and acoustic modeling in order to overcome the model inconsistencies between training speech and noisy input. In the two-stage recognition, the maximum a posteriori (MAP) based adaptation algorithm is used to incrementally adapt the noise model. In order to evaluate our proposed approach, a Mandarin keyword spotting system was constructed. The experimental results show our proposed method achieves a better recognition rate compared to the audible noise suppression (ANS) and parallel model combination (PMC) methods for both in 70km/hr (10.3dB) and 90km/hr (6.4dB) car environments.

Original language	English
Pages (from-to)	I/385-I/388
Journal	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	1
Publication status	Published - 2002 Jul 11
Event	2002 IEEE International Conference on Acustics, Speech, and Signal Processing - Orlando, FL, United States Duration: 2002 May 13 → 2002 May 17

All Science Journal Classification (ASJC) codes

Software
Signal Processing
Electrical and Electronic Engineering

Access to Document

Fingerprint Dive into the research topics of 'Perceptual speech modeling for noisy speech recognition'. Together they form a unique fingerprint.

Cite this

@article{bd0b1df883384634a90cb9e3b51082ef,

title = "Perceptual speech modeling for noisy speech recognition",

abstract = "This paper proposes a perceptual modeling approach with a two-stage recognition to deal with the issues of recognition degradation in noisy environment. The auditory masking effect is used for speech enhancement and acoustic modeling in order to overcome the model inconsistencies between training speech and noisy input. In the two-stage recognition, the maximum a posteriori (MAP) based adaptation algorithm is used to incrementally adapt the noise model. In order to evaluate our proposed approach, a Mandarin keyword spotting system was constructed. The experimental results show our proposed method achieves a better recognition rate compared to the audible noise suppression (ANS) and parallel model combination (PMC) methods for both in 70km/hr (10.3dB) and 90km/hr (6.4dB) car environments.",

author = "Wu, {Chung Hsien} and Chiu, {Yu Hsien} and Huigan Lim",

year = "2002",

month = jul,

day = "11",

language = "English",

volume = "1",

pages = "I/385--I/388",

journal = "Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing",

issn = "0736-7791",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

note = "2002 IEEE International Conference on Acustics, Speech, and Signal Processing ; Conference date: 13-05-2002 Through 17-05-2002",

}

TY - JOUR

T1 - Perceptual speech modeling for noisy speech recognition

AU - Wu, Chung Hsien

AU - Chiu, Yu Hsien

AU - Lim, Huigan

PY - 2002/7/11

Y1 - 2002/7/11

N2 - This paper proposes a perceptual modeling approach with a two-stage recognition to deal with the issues of recognition degradation in noisy environment. The auditory masking effect is used for speech enhancement and acoustic modeling in order to overcome the model inconsistencies between training speech and noisy input. In the two-stage recognition, the maximum a posteriori (MAP) based adaptation algorithm is used to incrementally adapt the noise model. In order to evaluate our proposed approach, a Mandarin keyword spotting system was constructed. The experimental results show our proposed method achieves a better recognition rate compared to the audible noise suppression (ANS) and parallel model combination (PMC) methods for both in 70km/hr (10.3dB) and 90km/hr (6.4dB) car environments.

AB - This paper proposes a perceptual modeling approach with a two-stage recognition to deal with the issues of recognition degradation in noisy environment. The auditory masking effect is used for speech enhancement and acoustic modeling in order to overcome the model inconsistencies between training speech and noisy input. In the two-stage recognition, the maximum a posteriori (MAP) based adaptation algorithm is used to incrementally adapt the noise model. In order to evaluate our proposed approach, a Mandarin keyword spotting system was constructed. The experimental results show our proposed method achieves a better recognition rate compared to the audible noise suppression (ANS) and parallel model combination (PMC) methods for both in 70km/hr (10.3dB) and 90km/hr (6.4dB) car environments.

UR - http://www.scopus.com/inward/record.url?scp=17444442112&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=17444442112&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:17444442112

VL - 1

SP - I/385-I/388

JO - Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing

JF - Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing

SN - 0736-7791

T2 - 2002 IEEE International Conference on Acustics, Speech, and Signal Processing

Y2 - 13 May 2002 through 17 May 2002

ER -