Advanced search
1 file | 195.25 KB Add to list

Exemplar-based joint channel and noise compensation

Author
Organization
Abstract
In this paper two models for channel estimation in exemplar-based noise robust speech recognition are proposed. Building on a compositional model that models noisy speech and a combination of noise and speech atoms, the first model iteratively estimates a filter to best compensate the mismatch with the observed noisy speech. The second model estimates separate filters for the noise and speech atoms. We show that both models enable noise-robust ASR even if the channel characteristics of the noisy speech do not match those of the exemplars in the dictionary. Moreover, the second model, which is able to estimate separate filters for speech and noise, is shown to be robust even in the presence of bandwidth-limited sources.
Keywords
noise robustness, channel compensation, matrix factorization, source separation, Speech recognition, SPEECH RECOGNITION

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 195.25 KB

Citation

Please use this url to cite or link to this publication:

MLA
Gemmeke, Jort F, Tuomas Virtanen, and Kris Demuynck. “Exemplar-based Joint Channel and Noise Compensation.” International Conference on Acoustics Speech and Signal Processing ICASSP. IEEE, 2013. 868–872. Print.
APA
Gemmeke, J. F., Virtanen, T., & Demuynck, K. (2013). Exemplar-based joint channel and noise compensation. International Conference on Acoustics Speech and Signal Processing ICASSP (pp. 868–872). Presented at the International Conference on Acoustics, Speech, and Signal Processing, IEEE.
Chicago author-date
Gemmeke, Jort F, Tuomas Virtanen, and Kris Demuynck. 2013. “Exemplar-based Joint Channel and Noise Compensation.” In International Conference on Acoustics Speech and Signal Processing ICASSP, 868–872. IEEE.
Chicago author-date (all authors)
Gemmeke, Jort F, Tuomas Virtanen, and Kris Demuynck. 2013. “Exemplar-based Joint Channel and Noise Compensation.” In International Conference on Acoustics Speech and Signal Processing ICASSP, 868–872. IEEE.
Vancouver
1.
Gemmeke JF, Virtanen T, Demuynck K. Exemplar-based joint channel and noise compensation. International Conference on Acoustics Speech and Signal Processing ICASSP. IEEE; 2013. p. 868–72.
IEEE
[1]
J. F. Gemmeke, T. Virtanen, and K. Demuynck, “Exemplar-based joint channel and noise compensation,” in International Conference on Acoustics Speech and Signal Processing ICASSP, Vancouver, Canada, 2013, pp. 868–872.
@inproceedings{4091792,
  abstract     = {In this paper two models for channel estimation in exemplar-based noise robust speech recognition are proposed. Building on a compositional model that models noisy speech and a combination of noise and speech atoms, the first model iteratively estimates a filter to best compensate the mismatch with the observed noisy speech. The second model estimates separate filters for the noise and speech atoms. We show that both models enable noise-robust ASR even if the channel characteristics of the noisy speech do not match those of the exemplars in the dictionary. Moreover, the second model, which is able to estimate separate filters for speech and noise, is shown to be robust even in the presence of bandwidth-limited sources.},
  author       = {Gemmeke, Jort F and Virtanen, Tuomas and Demuynck, Kris},
  booktitle    = {International Conference on Acoustics Speech and Signal Processing ICASSP},
  isbn         = {9781479903566},
  issn         = {1520-6149},
  keywords     = {noise robustness,channel compensation,matrix factorization,source separation,Speech recognition,SPEECH RECOGNITION},
  language     = {eng},
  location     = {Vancouver, Canada},
  pages        = {868--872},
  publisher    = {IEEE},
  title        = {Exemplar-based joint channel and noise compensation},
  year         = {2013},
}

Web of Science
Times cited: