Movatterモバイル変換

[0]ホーム
URL:

Skip to main content

Advertisement

Springer Nature Link
Log in

Vector Autoregressive Model for Missing Feature Reconstruction

  • Conference paper

Part of the book series:Lecture Notes in Computer Science ((LNAI,volume 4274))

Included in the following conference series:

  • 1636Accesses

Abstract

This paper proposes a Vector Autoregressive (VAR) model as a new technique for missing feature reconstruction in ASR. We model the spectral features using multiple VAR models. A VAR model predicts missing features as a linear function of a block of feature frames. We also propose two schemes for VAR training and testing. The experiments on AURORA-2 database have validated the modeling methodology and shown that the proposed schemes are especially effective for low SNR speech signals. The best setting has achieved a recognition accuracy of 88.2% at -5dB SNR on subway noise task when oracle data mask is used.

This is a preview of subscription content,log in via an institution to check access.

Access this chapter

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Similar content being viewed by others

References

  1. Raj, B., Stern, R.M.: Missing-feature approaches in speech reconitoin. Signal Processing Magazine 22(5), 101–116 (2005)

    Article  Google Scholar 

  2. Raj, B., Seltzer, M.L., Stern, R.M.: Reconstruction of missing features for robust speech recognition. Speech Communication 43(4), 275–296 (2004)

    Article  Google Scholar 

  3. Cooke, M., Green, P., Josifovski, L., Vizinho, A.: Robust automatic speech recognition with missing and uncertain acoustic data. Speech Communication 34, 267–285 (2001)

    Article MATH  Google Scholar 

  4. Lütkepohl, H.: Introduction to multiple time series analysis, 2nd edn. Springer, Heidelberg (1993)

    MATH  Google Scholar 

  5. Lim, J.S., Oppenheim, A.V.: Enhancement and bandwidth compression of noisy speech. Proceeding of the IEEE 67(12), 1586–1604 (1979)

    Article  Google Scholar 

  6. Van Hamme, H.: Robust speech recognition using cepstral domain missing data techniques and noisy mask. In: ICASSP 2004 (2004)

    Google Scholar 

  7. Pearce, D., Hirsch, H.-G.: The AURORA experimental framework for the performance evaluation of speech recogntion systems under noisy conditions. In: ICSLP, Beijing, China, pp. 29–32 (2000)

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. School of Computer Engineering, Nanyang Technological University, Singapore

    Xiong Xiao, Haizhou Li & Eng Siong Chng

  2. Institute for Infocomm Research, Singapore

    Xiong Xiao & Haizhou Li

Authors
  1. Xiong Xiao

    You can also search for this author inPubMed Google Scholar

  2. Haizhou Li

    You can also search for this author inPubMed Google Scholar

  3. Eng Siong Chng

    You can also search for this author inPubMed Google Scholar

Editor information

Editors and Affiliations

  1. Department of Computer Science, The University of Hong Kong, Hong Kong

    Qiang Huo

  2. Human Language Technology Department, Institute for Infocomm Research (I2R), 119613, Singapore

    Bin Ma

  3. School of Computer Engineering, Nanyang Technological University (NTU), 639798, Singapore

    Eng-Siong Chng

  4. Institute for Infocomm Research, 21 Heng Mui Keng Terrace, 119613, Singapore

    Haizhou Li

Rights and permissions

Copyright information

© 2006 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Xiao, X., Li, H., Chng, E.S. (2006). Vector Autoregressive Model for Missing Feature Reconstruction. In: Huo, Q., Ma, B., Chng, ES., Li, H. (eds) Chinese Spoken Language Processing. ISCSLP 2006. Lecture Notes in Computer Science(), vol 4274. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11939993_35

Download citation

Publish with us

[8]ページ先頭
©2009-2025 Movatter.jp