Advertisement

Small and Large Vocabulary Speech Recognition of MP3 Data under Real-Word Conditions: Experimental Study

  • Petr Pollak
  • Michal Borsky
Part of the Communications in Computer and Information Science book series (CCIS, volume 314)

Abstract

This paper presents the study of speech recognition accuracy both for small and large vocabulary task with respect to different levels of MP3 compression of processed data. The motivation behind the work was to evaluate the usage of ASR system for off-line automatic transcription of recordings collected from standard present MP3 devices under different levels of background noise and channel distortion. Although MP3 may not be an optimal compression algorithm, the performed experiments have prooved that it does not distort speech signal significantly for higher compression rates. Realized experiments showed also that the accuracy of speech recognition (both small- and large-vocabulary) decreased very slowly for the bit-rate of 24 kbps and higher. However, slightly different setup of speech feature computation is necessary for MP3 speech data, mainly PLP features give significantly better results in comparison to MFCC.

Keywords

Speech recognition Small vocabulary Large vocabulary LVCSR MPEG compression MP3 Noise robustness 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Nouza, J., Červa, P., Ždánský, J.: Very large vocabulary voice dictation for mobile devices. In: Proc. of Interspeech 2009, Brighton, UK, pp. 995–998 (2009)Google Scholar
  2. 2.
    Chen, S.S., Eide, E., Gales, M.J.F., Gopinath, R.A., Kanvesky, D., Olsen, P.: Automatic transcription of broadcast news. Speech Communication 37(1-2), 69–87 (2002)zbMATHCrossRefGoogle Scholar
  3. 3.
    Gauvain, J.-L., Lamel, L., Adda, G.: The LIMSI broadcast news transcription system. Speech Communication 37(1-2), 89–108 (2002)zbMATHCrossRefGoogle Scholar
  4. 4.
    Vaněk, J., Psutka, J.: Gender-dependent acoustic models fusion developed for automatic subtitling of parliament meetings broadcasted by the Czech TV. In: Proc. of Text, Speech and Dialog, Brno, pp. 431–438. Czech Republic (2010)Google Scholar
  5. 5.
    Psutka, J., Psutka, J., Ircing, P., Hoidekr, J.: Recognition of spontaneously pronounced TV ice-hockey commentary. In: Proc. of ISCA & IEEE Workshop on Spontaneous Speech Processing and Recognition, Tokyo, pp. 83–86 (2003)Google Scholar
  6. 6.
    Makhoul, J., Kubala, F., Leek, T., Liu, D., Nguyen, L., Schwartz, R., Srivastava, A.: Speech and language technologies for audio indexing and retrieval. Proc. of the IEEE 88(8), 1338–1353 (2000)CrossRefGoogle Scholar
  7. 7.
    Byrne, W., Doermann, D., Franz, M., Gustman, S., Hajič, J., Oard, D., Pichney, M., Psutka, J., Ramabhadran, B., Soergel, D., Ward, T., Zhu, W.J.: Automatic recognition of spontaneous speech for access to multilingual oral history archives. IEEE Trans. on Speech and Audio Processing 12(4), 420–435 (2004)CrossRefGoogle Scholar
  8. 8.
    Bouvigne, G.: MP3 standard. Homepage (2007), http://www.mp3-tech.org
  9. 9.
    Brandenburg, K., Popp, H.: An introduction to MPEG layer 3. EBU Technical Review (June 2000)Google Scholar
  10. 10.
    Barras, C., Lamel, L., Gauvain, J.L.: Automatic transcription of compressed broadcast audio. In: Proc. of the IEEE International Conference on Acoustics, Speech, and Signal Processing, Salt Lake City, USA, pp. 265–268 (2001)Google Scholar
  11. 11.
    Pollak, P., Behunek, M.: Accuracy of MP3 speech recognition under real-world conditions. Experimental study. In: Proc. of SIGMAP 2011 - International Conference on Signal Processing and Multimedia Applications, Seville, Spain, vol. 1, pp. 5–10 (July 2011)Google Scholar
  12. 12.
    Fousek, P., Pollák, P.: Additive noise and channel distortion-robust parameterization tool. performance evaluation on Aurora 2 & 3. In: Proc. of Eurospeech 2003, Geneve, Switzerland (2003)Google Scholar
  13. 13.
    Bořil, H., Fousek, P., Pollák, P.: Data-driven design of front-end filter bank for Lombard speech recognition. In: Proc. of ICSLP 2006, Pittsburgh, USA (2006)Google Scholar
  14. 14.
    Rajnoha, J., Pollák, P.: ASR systems in noisy environment: Analysis and solutions for increasing noise robustness. Radioengineering 20(1), 74–84 (2011)Google Scholar
  15. 15.
    ITU-T: International Telecommunication Union Recommendation G.729, coding of speech at 8 kbit/s using conjugate-structure algebraic-code-excited linear prediction(CS-ACELP) (2007), http://www.itu.int/ITU-T
  16. 16.
    ETSI: Digital cellular telecommunications system (Phase 2+) (GSM). Test sequences for the Adaptive Multi-Rate (AMR) speech codec (2007), http://www.etsi.org
  17. 17.
    Valin, J.M.: The speex codec manual. version 1.2 beta 3 (2007), http://www.speex.org
  18. 18.
    Cheng, M., et. al.: LAME MP3 encoder 3.99 alpha 10 (2008), http://www.free-codecs.com
  19. 19.
    Huang, X., Acero, A., Hon, H.-W.: Spoken Language Processing. Prentice Hall (2001)Google Scholar
  20. 20.
    Young, S., et al.: The HTK Book, Version 3.4.1, Cambridge (2009)Google Scholar
  21. 21.
    Psutka, J., Müller, L., Psutka, J.V.: Comparison of MFCC and PLP parameterization in the speaker independent continuous speech recognition task. In: Proc. of Eurospeech 2001, Aalborg, Denmark (2001)Google Scholar
  22. 22.
    Hermansky, H.: Perceptual linear predictive (PLP) analysis of speech. Journal of the Acoustical Society of America 87(4), 1738–1752 (1990)CrossRefGoogle Scholar
  23. 23.
    Institute of the Czech National Corpus: Homepage (2010) http://www.korpus.cz.
  24. 24.
    Institute of the Czech National Corpus: SYN2006PUB - corpus of newspapers and magazines from 1989 - 2004 (2006), http://ucnk.ff.cuni.cz/english/syn2006pub.php
  25. 25.
    Prochazka, V., Pollak, P., Zdansky, J., Nouza, J.: Performance of Czech speech recognition with language models created from public resources. Radioengineering 20(4), 1002–1008 (2011)Google Scholar
  26. 26.
    Fousek, P.: CtuCopy-Universal feature extractor and speech enhancer (2006), http://noel.feld.cvut.cz/speechlab
  27. 27.
    ELRA: Czech SPEECON database. Catalog No. S0298 (2009), http://www.elra.info
  28. 28.
    Pollák, P., Černocký, J.: Czech SPEECON adult database. Technical report (April 2004)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Petr Pollak
    • 1
  • Michal Borsky
    • 1
  1. 1.Faculty of Electrical EngineeringCzech Technical University in PraguePragueCzech Republic

Personalised recommendations