EMG signal processing for audio-emg-based multi-modal speech recognition

Zhipeng Zhang, Hiroyuki Manabe, Tsutomu Horikoshi, Tomoyuki Ohya

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

1 Citation (Scopus)

Abstract

This paper proposes robust methods for processing EMG (electromyography) signals in the framework of audio-EMG-based speech recognition. The EMG signals are captured when uttered and used as auxiliary information for recognizing speech. Two robust methods (Cepstral Mean Normalization and Spectral Subtraction) for EMG signal processing are investigated to improve the recognition performance. We also investigate the importance of stream weighting in audio-EMG-based multi-modal speech recognition. Experiments are carried out at various noise conditions and the results show the effectiveness of the proposed methods. A significant improvement in word accuracy over the audio-only recognition scheme is achieved by combining the methods.

Original languageEnglish
Title of host publicationProceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005
EditorsM.H. Hamza
Pages430-433
Number of pages4
Publication statusPublished - 2005 Nov 28
Externally publishedYes
Event3rd IASTED International Conference on Medical Engineering 2005 - Innsbruck, Austria
Duration: 2005 Feb 162005 Feb 18

Publication series

NameProceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005

Conference

Conference3rd IASTED International Conference on Medical Engineering 2005
CountryAustria
CityInnsbruck
Period05/2/1605/2/18

Fingerprint

Electromyography
Speech recognition
Signal processing
Processing
Experiments

Keywords

  • Cepstral Mean Subtraction
  • EMG
  • Multimodal speech recognition
  • Spectral Subtraction

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Zhang, Z., Manabe, H., Horikoshi, T., & Ohya, T. (2005). EMG signal processing for audio-emg-based multi-modal speech recognition. In M. H. Hamza (Ed.), Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005 (pp. 430-433). (Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005).

EMG signal processing for audio-emg-based multi-modal speech recognition. / Zhang, Zhipeng; Manabe, Hiroyuki; Horikoshi, Tsutomu; Ohya, Tomoyuki.

Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005. ed. / M.H. Hamza. 2005. p. 430-433 (Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005).

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

Zhang, Z, Manabe, H, Horikoshi, T & Ohya, T 2005, EMG signal processing for audio-emg-based multi-modal speech recognition. in MH Hamza (ed.), Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005. Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005, pp. 430-433, 3rd IASTED International Conference on Medical Engineering 2005, Innsbruck, Austria, 05/2/16.
Zhang Z, Manabe H, Horikoshi T, Ohya T. EMG signal processing for audio-emg-based multi-modal speech recognition. In Hamza MH, editor, Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005. 2005. p. 430-433. (Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005).
Zhang, Zhipeng ; Manabe, Hiroyuki ; Horikoshi, Tsutomu ; Ohya, Tomoyuki. / EMG signal processing for audio-emg-based multi-modal speech recognition. Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005. editor / M.H. Hamza. 2005. pp. 430-433 (Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005).
@inproceedings{9b4afa8c709a4066bdaa762583b9e6e9,
title = "EMG signal processing for audio-emg-based multi-modal speech recognition",
abstract = "This paper proposes robust methods for processing EMG (electromyography) signals in the framework of audio-EMG-based speech recognition. The EMG signals are captured when uttered and used as auxiliary information for recognizing speech. Two robust methods (Cepstral Mean Normalization and Spectral Subtraction) for EMG signal processing are investigated to improve the recognition performance. We also investigate the importance of stream weighting in audio-EMG-based multi-modal speech recognition. Experiments are carried out at various noise conditions and the results show the effectiveness of the proposed methods. A significant improvement in word accuracy over the audio-only recognition scheme is achieved by combining the methods.",
keywords = "Cepstral Mean Subtraction, EMG, Multimodal speech recognition, Spectral Subtraction",
author = "Zhipeng Zhang and Hiroyuki Manabe and Tsutomu Horikoshi and Tomoyuki Ohya",
year = "2005",
month = "11",
day = "28",
language = "English",
isbn = "0889864780",
series = "Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005",
pages = "430--433",
editor = "M.H. Hamza",
booktitle = "Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005",

}

TY - GEN

T1 - EMG signal processing for audio-emg-based multi-modal speech recognition

AU - Zhang, Zhipeng

AU - Manabe, Hiroyuki

AU - Horikoshi, Tsutomu

AU - Ohya, Tomoyuki

PY - 2005/11/28

Y1 - 2005/11/28

N2 - This paper proposes robust methods for processing EMG (electromyography) signals in the framework of audio-EMG-based speech recognition. The EMG signals are captured when uttered and used as auxiliary information for recognizing speech. Two robust methods (Cepstral Mean Normalization and Spectral Subtraction) for EMG signal processing are investigated to improve the recognition performance. We also investigate the importance of stream weighting in audio-EMG-based multi-modal speech recognition. Experiments are carried out at various noise conditions and the results show the effectiveness of the proposed methods. A significant improvement in word accuracy over the audio-only recognition scheme is achieved by combining the methods.

AB - This paper proposes robust methods for processing EMG (electromyography) signals in the framework of audio-EMG-based speech recognition. The EMG signals are captured when uttered and used as auxiliary information for recognizing speech. Two robust methods (Cepstral Mean Normalization and Spectral Subtraction) for EMG signal processing are investigated to improve the recognition performance. We also investigate the importance of stream weighting in audio-EMG-based multi-modal speech recognition. Experiments are carried out at various noise conditions and the results show the effectiveness of the proposed methods. A significant improvement in word accuracy over the audio-only recognition scheme is achieved by combining the methods.

KW - Cepstral Mean Subtraction

KW - EMG

KW - Multimodal speech recognition

KW - Spectral Subtraction

UR - http://www.scopus.com/inward/record.url?scp=27844589598&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27844589598&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:27844589598

SN - 0889864780

T3 - Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005

SP - 430

EP - 433

BT - Proceedings of the 3rd IASTED International Conference on Biomedical Engineering 2005

A2 - Hamza, M.H.

ER -