Skip to main content
SpringerLink
Log in

Deep Neural Network Based Continuous Speech Recognition for Serbian Using the Kaldi Toolkit

  • Conference paper
  • First Online:
Speech and Computer (SPECOM 2015)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 9319))

Included in the following conference series:

Abstract

This paper presents a deep neural network (DNN) based large vocabulary continuous speech recognition (LVCSR) system for Serbian, developed using the open-source Kaldi speech recognition toolkit. The DNNs are initialized using stacked restricted Boltzmann machines (RBMs) and trained using cross-entropy as the objective function and the standard error backpropagation procedure in order to provide posterior probability estimates for the hidden Markov model (HMM) states. Emission densities of HMM states are represented as Gaussian mixture models (GMMs). The recipes were modified based on the particularities of the Serbian language in order to achieve the optimal results. A corpus of approximately 90 hours of speech (21000 utterances) is used for the training. The performances are compared for two different sets of utterances between the baseline GMM-HMM algorithm and various DNN settings.

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

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 39.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 54.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

References

  1. Delić, V., Sečujski, M., Jakovljević, N., Pekar, D., Mišković, D., Popović, B., Ostrogonac, S., Bojanić, M., Knežević, D.: Speech and language resources within speech recognition and synthesis systems for Serbian and kindred south slavic languages. In: Železný, M., Habernal, I., Ronzhin, A. (eds.) SPECOM 2013. LNCS, vol. 8113, pp. 319–326. Springer, Heidelberg (2013)

    Chapter  Google Scholar 

  2. Young, S.J., Odell, J., Woodland, P.C.: Tree-based state tying for high accuracy acoustic modelling. In: ARPA Human Language Technology Workshop, pp. 307–312, Princeton (1994)

    Google Scholar 

  3. Jakovljević, N., Mišković, D., Janev, M., Pekar, D.: A decoder for large vocabulary speech recognition. In: 18th International Conference on Systems, Signals and Image Processing, IWSSIP, pp. 1–4, Sarajevo (2011)

    Google Scholar 

  4. Povey, D., et al.: The Kaldi speech recognition toolkit. In: IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU, pp. 1–4, Waikoloa (2011)

    Google Scholar 

  5. Mohri, M., Pereira, F., Riley, M.: Weighted finite-state transducers in speech recognition. Comput. Speech Lang. 16, 69–88 (2002)

    Article  Google Scholar 

  6. Blackford, L.S., et al.: An updated set of basic linear algebra subprograms (BLAS). ACM Trans. Math. Softw. 28(2), 135–151 (2002)

    Article  MathSciNet  Google Scholar 

  7. Anderson, E., et al.: LAPACK Users’ Guide. Society for Industrial and Applied Mathematics (SIAM), Philadelphia (1999)

    Book  Google Scholar 

  8. Popović, B., Pakoci, E., Ostrogonac, S., Pekar, D.: Large vocabulary continuous speech recognition for Serbian using the Kaldi toolkit. In: 10th Digital Speech and Image Processing, DOGS, pp. 31–34, Novi Sad (2014)

    Google Scholar 

  9. Veselý, K., Arnab, G., Lukáš, B., Povey, D.: Sequence-discriminative training of deep neural networks. In: International Speech Communication Association, Interspeech 2013, pp. 2345–2349, Lyon (2013)

    Google Scholar 

  10. Povey, D., Kanevsky, D., Kingsbury, B., Ramabhadran, B., Saon, G., Visweswariah, K.: Boosted MMI for model and feature-space discriminative training. In: 33rd International Conference on Acoustics, Speech and Signal Processing, ICASSP, pp. 4057–4060, Las Vegas (2008)

    Google Scholar 

  11. Povey D., Woodland, P.C.: Minimum phone error and i-smoothing for improved discriminative training. In: 27th International Conference on Acoustics, Speech and Signal Processing, ICASSP, pp. I-105–I-108, Orlando (2002)

    Google Scholar 

  12. Povey, D., Kuo, H-K.J., Soltau, H.: Fast speaker adaptive training for speech recognition. In: 9th Annual Conference of the International Speech Communication Association, INTERSPEECH, pp. 1245–1248, Brisbane (2008)

    Google Scholar 

  13. Povey, D., et al.: The subspace Gaussian mixture model - a structured model for speech recognition. Comput. Speech Lang. 25, 404–439 (2011)

    Article  Google Scholar 

  14. Carreira-Perpiñán, M., Hinton, G.: On contrastive divergence learning. In: 10th International Workshop on Artifitial Intelligence and Statistic, AISTATS, pp. 59–66, Barbados (2005)

    Google Scholar 

  15. Stolcke, A., Zheng, J., Wang, W., Abrash, V.: SRILM at sixteen: update and outlook. In: IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU, Waikoloa (2011)

    Google Scholar 

  16. Kneser, R., Ney, H.: Improved backing-off for M-gram language modeling. In: 20th International Conference on Acoustics, Speech and Signal Processing, ICASSP, pp. 181–184, Detroit (1995)

    Google Scholar 

Download references

Acknowledgments

The work described in this paper was supported in part by the Ministry of Education, Science and Technological Development of the Republic of Serbia, within the project TR32035: “Development of Dialogue Systems for Serbian and Other South Slavic Languages”.

Author information

Authors and Affiliations

  1. Faculty of Technical Sciences, University of Novi Sad, Novi Sad, Serbia

    Branislav Popović, Stevan Ostrogonac, Edvin Pakoci, Nikša Jakovljević & Vlado Delić

Authors
  1. Branislav Popović
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Stevan Ostrogonac
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Edvin Pakoci
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Nikša Jakovljević
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Vlado Delić
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Branislav Popović .

Editor information

Editors and Affiliations

  1. SPIIRAS, Saint-Petersburg, Russia

    Andrey Ronzhin

  2. Moscow State Linguistic University, Moscow, Russia

    Rodmonga Potapova

  3. University of Patras, Patras, Greece

    Nikos Fakotakis

Rights and permissions

Reprints and permissions

Copyright information

© 2015 Springer International Publishing Switzerland

About this paper

Cite this paper

Popović, B., Ostrogonac, S., Pakoci, E., Jakovljević, N., Delić, V. (2015). Deep Neural Network Based Continuous Speech Recognition for Serbian Using the Kaldi Toolkit. In: Ronzhin, A., Potapova, R., Fakotakis, N. (eds) Speech and Computer. SPECOM 2015. Lecture Notes in Computer Science(), vol 9319. Springer, Cham. https://doi.org/10.1007/978-3-319-23132-7_23

Download citation

  • DOI: https://doi.org/10.1007/978-3-319-23132-7_23

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-319-23131-0

  • Online ISBN: 978-3-319-23132-7

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation