Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
Springer Nature Link
Log in

Integration of multiple acoustic and language models for improved Hindi speech recognition system

  • Published:
International Journal of Speech Technology Aims and scope Submit manuscript

Abstract

Despite the significant progress of automatic speech recognition (ASR) in the past three decades, it could not gain the level of human performance, particularly in the adverse conditions. To improve the performance of ASR, various approaches have been studied, which differ in feature extraction method, classification method, and training algorithms. Different approaches often utilize complementary information; therefore, to use their combination can be a better option. In this paper, we have proposed a novel approach to use the best characteristics of conventional, hybrid and segmental HMM by integrating them with the help of ROVER system combination technique. In the proposed framework, three different recognizers are created and combined, each having its own feature set and classification technique. For design and development of the complete system, three separate acoustic models are used with three different feature sets and two language models. Experimental result shows that word error rate (WER) can be reduced about 4% using the proposed technique as compared to conventional methods. Various modules are implemented and tested for Hindi Language ASR, in typical field conditions as well as in noisy environment.

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

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8

Similar content being viewed by others

Explore related subjects

Discover the latest articles, news and stories from top researchers in related subjects.

References

  • Aggarwal, R. K., & Dave, M. (2011a). Acoustic modeling problem for automatic speech recognition system: conventional methods (Part I). International Journal of Speech Technology, 14, 297–308.

    Article  Google Scholar 

  • Aggarwal, R. K., & Dave, M. (2011b). Acoustic modeling problem for automatic speech recognition system: advances and refinements (Part II). International Journal of Speech Technology, 14, 309–320.

    Article  Google Scholar 

  • Aubert, X. L. (2002). An overview of decoding techniques for large vocabulary continuous speech recognition. Computer Speech and Language, 16(1), 89–114.

    Article  Google Scholar 

  • Becchetti, C., & Ricotti, K. P. (2004). Speech recognition theory and C++ implementation. New York: Wiley.

    Google Scholar 

  • Benouareth, A., Ennaji, A., & Sellami, M. (2008). Semi-continuous HMMs with explicit state duration for unconstrained Arabic word modeling and recognition. Pattern Recognition Letters, 29, 1742–1752.

    Article  Google Scholar 

  • Beyerlein, P. (1998). Discriminative model combination. In Proceedings ICASSP (pp. 481–484).

    Google Scholar 

  • Bourlard, H., Morgan, N., & Renals, S. (1992). Neural nets and hidden Markov models: review and generalizations. Speech Communication, 11, 237–246.

    Article  Google Scholar 

  • Chopde, A. (2009). Itrans Indian language transliteration package version, 5.2 source. http://www.aczone.com/itrans/.

  • Davis, S., & Mermelstein, P. (1980). Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences. IEEE Transactions on Acoustics, Speech, and Signal Processing, 28, 357–366.

    Article  Google Scholar 

  • Digalakis, V. V., & Murveit, H. (1994). Genones: Optimization the degree of tying in a large vocabulary HMM-based speech recognizer. In Proceeding of IEEE ICASSP (pp. 537–540).

    Google Scholar 

  • Fiscus, J. (1997). A post-processing system to yield reduced word error rates: recognizer output voting error reduction (ROVER). In Proceeding of the IEEE workshop on automatic speech recognition and understanding (ASRU’97), Santa Barbara (pp. 347–352).

    Chapter  Google Scholar 

  • Garau, G., & Renals, S. (2008). Combining spectral representations for large-vocabulary continuous speech recognition. IEEE Transactions on Audio, Speech, and Language Processing, 16(3), 508–518.

    Article  Google Scholar 

  • Haeb-Umbach, R., & Ney, H. (1992). Linear discriminant analysis for improved large vocabulary continuous speech recognition. In Proceedings of ICASSP (pp. 13–16).

    Google Scholar 

  • Hagen, A., & Morris, A. (2005). Recent advances in the multi-stream HMM/ANN hybrid approach to noise robust ASR. Computer Speech and Language, 19(3), 3–30.

    Article  Google Scholar 

  • Hermansky, H. (1990). Perceptually predictive (PLP) analysis of speech. The Journal of the Acoustical Society of America, 87, 1738–1752.

    Article  Google Scholar 

  • Kirchhoff, K., Fink, G. A., & Sagerer, G. (2000). Conversational speech recognition using acoustic and articulatory input. In Proceeding IEEE ICASSP, Istanbul, Turkey.

    Google Scholar 

  • Kirchhoff, K., & Bilmes, J. A. (2000). Combination and joint Training of acoustic classifiers for speech Recognition. In ISCA ITRW workshop on automatic speech recognition: challenges for the new mllennium, Paris, France.

    Google Scholar 

  • Kumar, N., & Andreou, A. G. (1998). Heteroscedastic discriminant analysis and reduced rank HMMs for improved speech recognition. Speech Communication, 26, 283–297.

    Article  Google Scholar 

  • Leggetter, C. J., & Woodland, P. (1995). Speaker adaptation using maximum likelihood linear regression. Computer Speech and Language, 9(2), 171–185.

    Article  Google Scholar 

  • Mangu, L., Brill, E., & Stolcke, A. (2000). Finding consensus in speech recognition: word error minimization and other application of confusion network. Computer Speech and Language, 14(4), 373–400.

    Article  Google Scholar 

  • O’Shaughnessy, D. (2008). Automatic speech recognition: history, methods and challenges. Pattern Recognition, 41, 2965–2979. Invited paper.

    Article  MATH  Google Scholar 

  • Padmanabhan, M., & Picheny, M. (2002). Large vocabulary speech recognition algorithms. IEEE Computer Society, 35(4), 42–50.

    Article  Google Scholar 

  • Rao, G. V. R., & Yegnanarayana, B. (1991). Word boundary hypothesization in Hindi speech. Computer Speech and Language, 5, 379–392.

    Article  Google Scholar 

  • Rao, K. S. (2011). Application of prosody models for developing speech systems in Indian languages. International Journal of Speech Technology, 14, 19–33.

    Article  Google Scholar 

  • Samir, A., Duchateau, J., & Vanhamme, H. (2008). Discriminative model combination and language model selection in a reading tutor for children. In Interspeech, ISCA, Brisbane Australia (pp. 2795–2798).

    Google Scholar 

  • Sankar, A. (2005). Bayesian model combination (Baycom) for improved recognition. In Proceedings IEEE international conference on acoustics, speech, and signal processing.

    Google Scholar 

  • Saraswathi, S., & Geetha, T. (2007). Comparison of morpheme-based language model with different word-based models for improving the performance of Tamil speech recognition system. ACM Transaction on Asian Language Information Processing, 6(3), article 9.

  • Schwenk, H., & Gauvain, J.-L. (2000). Combining multiple speech recognizers using voting and language model information. In IEEE international conference on spoken language processing (ICSLP), II Pekin (pp. 915–918).

    Google Scholar 

  • Silsbee, P., & Bovik, A. (1996). Computer lip-reading for improved accuracy in ASR. IEEE Transactions on Speech and Audio Processing, 4(5), 337–351.

    Article  Google Scholar 

  • Siohan, O., Ramabhadran, B., & Kingsbury, B. (2005). Constructing ensembles of ASR systems using randomized decision trees. In ICASSP (Vol. I, pp. 197–200).

    Google Scholar 

  • Stolcke, A., et al. (2000). The SRI March 2000 Hub-5 conversational speech transcription system. In Proc. speech transcription workshop.

    Google Scholar 

  • Stolke, A., Konig, Y., & Weintraub, M. (1997). Explicit word error minimization in N-best list rescoring. In Proc. Eurospeech (Vol. 1, pp. 163–166).

    Google Scholar 

  • Vaid, J., & Gupta, A. (2002). Exploring word recognition in a semi alphabetic script: the case of devnagari. Brain and Language, 81, 679–690.

    Article  Google Scholar 

  • Varga, A., & Steeneken, H. J. M. (1993). Assessment for automatic recognition: II. NOISEX-92: a database and an experiment to study the effect of additive noise on speech recognition systems. ESCA Journal of Speech Communication, 12(3), 247–251.

    Article  Google Scholar 

  • Waibel, A., Sawai, H., & Shikano, K. (1989). Modularity and scaling in large phonemic neural networks. IEEE Transactions on Acoustics, Speech, and Signal Processing, 37(12), 1888–1898.

    Article  Google Scholar 

  • Woodland, P., Gales, M., Pye, D., & Young, S. (1997). Broadcast news transcription using HTK. In Proceeding of IEEE international conference on acoustics, speech and signal processing, ICASSP, Munich, Germany (Vol. 2, pp. 719–722).

    Google Scholar 

  • Young, S., et al. (2009). The HTK Book. Microsoft Corporation and Cambridge University Engineering Department.

  • Zhang, R., & Rudnicky, A. (2006). Investigations of Issues for using multiple acoustic models to improve CSR. In IEEE international conference on spoken language processing, Pitsburgh, PA, USA.

    Google Scholar 

  • Zolney, A., Kocharov, D., Schluter, R., & Ney, H. (2007). Using multiple acoustic feature sets for speech recognition. Speech Communication, 49, 514–525.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Computer Engineering, N.I.T., Kurukshetra, India

    R. K. Aggarwal & M. Dave

Authors
  1. R. K. Aggarwal
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. M. Dave
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to R. K. Aggarwal.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Aggarwal, R.K., Dave, M. Integration of multiple acoustic and language models for improved Hindi speech recognition system. Int J Speech Technol 15, 165–180 (2012). https://doi.org/10.1007/s10772-012-9131-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10772-012-9131-y

Keywords

Search

Navigation