Recognition of Marathi Numerals Using MFCC and DTW Features

  • Siddheshwar S. GangondaEmail author
  • Prashant P. Patavardhan
  • Kailash J. Karande
Conference paper
Part of the Communications in Computer and Information Science book series (CCIS, volume 1037)


Numeral recognition is one amongst the foremost important problems in pattern recognition. Its numerous uses like reading communication postal code, worker code, bank cheque method etc. To the simplest of our information, very less work has been wiped out Marathi language as compared with other Indian and non-Indian languages. It has mentioned a unique technique for recognition of isolated Marathi numerals. It introduces Marathi numerals and identification technique using MFCC and DTW as attributes. The accuracy of the pre-recorded samples is greater than that of online testing samples. We have got additionally seen that the accuracy of the speaker dependent samples is over that of the speaker independent samples. Another technique known as HMM is additionally discussed. By experimentation, it’s ascertained that identification exactness is higher for HMM than DTW, but the training method in DTW is extremely straightforward and quick, as compared to HMM. The time needed for recognition of numerals using HMM is additional as compared to DTW, because it should bear the various states, iterations and lots of additional mathematical modeling, thus DTW is most well-liked for the real-time applications.


Hidden Markov Model (HMM) Mel-Frequency Cepstral Coefficient (MFCC) Distance Time Warping (DTW) 


  1. 1.
    Saon, G., Picheny, M.: Recent advances in conversational speech recognition using convolutional and recurrent neural networks. IBM J. Res. Dev. 61(4/5), 1 (2017)CrossRefGoogle Scholar
  2. 2.
    O’Shaughnessy, D.: Automatic speech recognition. In: CHILEAN Conference on Electrical, Electronics Engineering, Information and Communication Technologies (CHILECON) (2015).
  3. 3.
    Rabiner, L.R., Schafer, R.W.: Digital Processing Of Speech Signals, Low Price Edition. Pearson, London (2007)Google Scholar
  4. 4.
    Anusuya, M.A., Katti, S.K.: Speech recognition by machine: a review. (IJCSIS) Int. J. Comput. Sci. Inf. Secur. 6(3), 181–205 (2009)Google Scholar
  5. 5.
    Gawali, B.W., Gaikwad, S., Yannawar, P., Mehrotra, S.C.: Marathi isolated word recognition system using MFCC and DTW features. In: Proceedings of International Conference of Advances in Computer Science (2010)Google Scholar
  6. 6.
    Muda, L., Begam, M., Elamvazuth, I.: Voice recognition algorithms using Mel Frequency Cepstral Coefficient (MFCC) and Dynamic Time Warping (DTW) techniques. J. Comput. 2(3) (2010). ISSN 2151-9617Google Scholar
  7. 7.
    Vimala, C., Radhab, V.: Speaker independent isolated speech recognition system for Tamil language using HMM. In: International Conference on Communication Technology and System Design (2011)Google Scholar
  8. 8.
    Bala, A., Kumar, A., Birla, N.: Voice command recognition system based on MFCC and DTW. Int. J. Eng. Sci. Technol. 2(12), 7335–7342 (2010)Google Scholar
  9. 9.
    Jiang, H., Li, X., Liu, C.: Large margin hidden Markov models for speech recognition. IEEE Trans. Audio Speech Lang. Process. 14(5) (2006)Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  • Siddheshwar S. Gangonda
    • 1
    Email author
  • Prashant P. Patavardhan
    • 2
  • Kailash J. Karande
    • 1
  1. 1.SKN Sinhgad COEPandharpurIndia
  2. 2.KLS Gogte Institute of TechnologyBelagaviIndia

Personalised recommendations