PCMM-based feature compensation schemes using model interpolation and mixture sharing

Wooil Kim, Ohil Kwon, Hanseok Ko

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

In this paper, we propose an effective feature compensation scheme based on the speech model in order to achieve robust speech recognition. The proposed feature compensation method is based on parallel combined mixture model (PCMM). The previous PCMM works require a highly sophisticated procedure for estimation of the combined mixture model in order to reflect the time-varying noisy conditions at every utterance. The proposed schemes can cope with the time-varying background noise by employing the interpolation method of the multiple mixture models. We apply the 'data-driven' method to PCMM for more reliable model combination and introduce a frame-synched version for estimation of environments posteriori. In order to reduce the computational complexity due to multiple models, we propose a technique for mixture sharing. The statistically similar Gaussian components are selected and the smoothed versions are generated for sharing. The performance was examined over Aurora 2. 0 and speech corpus recorded while car-driving. The experimental results indicate that the proposed schemes are effective in realizing robust speech recognition and reducing the computational complexities under both simulated environments and real-life conditions.

Original languageEnglish
Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume1
Publication statusPublished - 2004
EventProceedings - IEEE International Conference on Acoustics, Speech, and Signal Processing - Montreal, Que, Canada
Duration: 2004 May 172004 May 21

Other

OtherProceedings - IEEE International Conference on Acoustics, Speech, and Signal Processing
CountryCanada
CityMontreal, Que
Period04/5/1704/5/21

Fingerprint

interpolation
Interpolation
speech recognition
Speech recognition
Computational complexity
Compensation and Redress
background noise
Railroad cars

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Acoustics and Ultrasonics

Cite this

Kim, W., Kwon, O., & Ko, H. (2004). PCMM-based feature compensation schemes using model interpolation and mixture sharing. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (Vol. 1)

PCMM-based feature compensation schemes using model interpolation and mixture sharing. / Kim, Wooil; Kwon, Ohil; Ko, Hanseok.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. Vol. 1 2004.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kim, W, Kwon, O & Ko, H 2004, PCMM-based feature compensation schemes using model interpolation and mixture sharing. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. vol. 1, Proceedings - IEEE International Conference on Acoustics, Speech, and Signal Processing, Montreal, Que, Canada, 04/5/17.
Kim W, Kwon O, Ko H. PCMM-based feature compensation schemes using model interpolation and mixture sharing. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. Vol. 1. 2004
Kim, Wooil ; Kwon, Ohil ; Ko, Hanseok. / PCMM-based feature compensation schemes using model interpolation and mixture sharing. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. Vol. 1 2004.
@inproceedings{39a1b73480014ca19cecb82ab434baea,
title = "PCMM-based feature compensation schemes using model interpolation and mixture sharing",
abstract = "In this paper, we propose an effective feature compensation scheme based on the speech model in order to achieve robust speech recognition. The proposed feature compensation method is based on parallel combined mixture model (PCMM). The previous PCMM works require a highly sophisticated procedure for estimation of the combined mixture model in order to reflect the time-varying noisy conditions at every utterance. The proposed schemes can cope with the time-varying background noise by employing the interpolation method of the multiple mixture models. We apply the 'data-driven' method to PCMM for more reliable model combination and introduce a frame-synched version for estimation of environments posteriori. In order to reduce the computational complexity due to multiple models, we propose a technique for mixture sharing. The statistically similar Gaussian components are selected and the smoothed versions are generated for sharing. The performance was examined over Aurora 2. 0 and speech corpus recorded while car-driving. The experimental results indicate that the proposed schemes are effective in realizing robust speech recognition and reducing the computational complexities under both simulated environments and real-life conditions.",
author = "Wooil Kim and Ohil Kwon and Hanseok Ko",
year = "2004",
language = "English",
volume = "1",
booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

}

TY - GEN

T1 - PCMM-based feature compensation schemes using model interpolation and mixture sharing

AU - Kim, Wooil

AU - Kwon, Ohil

AU - Ko, Hanseok

PY - 2004

Y1 - 2004

N2 - In this paper, we propose an effective feature compensation scheme based on the speech model in order to achieve robust speech recognition. The proposed feature compensation method is based on parallel combined mixture model (PCMM). The previous PCMM works require a highly sophisticated procedure for estimation of the combined mixture model in order to reflect the time-varying noisy conditions at every utterance. The proposed schemes can cope with the time-varying background noise by employing the interpolation method of the multiple mixture models. We apply the 'data-driven' method to PCMM for more reliable model combination and introduce a frame-synched version for estimation of environments posteriori. In order to reduce the computational complexity due to multiple models, we propose a technique for mixture sharing. The statistically similar Gaussian components are selected and the smoothed versions are generated for sharing. The performance was examined over Aurora 2. 0 and speech corpus recorded while car-driving. The experimental results indicate that the proposed schemes are effective in realizing robust speech recognition and reducing the computational complexities under both simulated environments and real-life conditions.

AB - In this paper, we propose an effective feature compensation scheme based on the speech model in order to achieve robust speech recognition. The proposed feature compensation method is based on parallel combined mixture model (PCMM). The previous PCMM works require a highly sophisticated procedure for estimation of the combined mixture model in order to reflect the time-varying noisy conditions at every utterance. The proposed schemes can cope with the time-varying background noise by employing the interpolation method of the multiple mixture models. We apply the 'data-driven' method to PCMM for more reliable model combination and introduce a frame-synched version for estimation of environments posteriori. In order to reduce the computational complexity due to multiple models, we propose a technique for mixture sharing. The statistically similar Gaussian components are selected and the smoothed versions are generated for sharing. The performance was examined over Aurora 2. 0 and speech corpus recorded while car-driving. The experimental results indicate that the proposed schemes are effective in realizing robust speech recognition and reducing the computational complexities under both simulated environments and real-life conditions.

UR - http://www.scopus.com/inward/record.url?scp=4544324741&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4544324741&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:4544324741

VL - 1

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

ER -