Spoken Language Identification for Indian Languages Using Split and Merge EM Algorithm

  • Naresh Manwani
  • Suman K. Mitra
  • M. V. Joshi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4815)


Performance of Language Identification (LID) System using Gaussian Mixture Models (GMM) is limited by the convergence of Expectation Maximization (EM) algorithm to local maxima. In this paper an LID system is described using Gaussian Mixture Models for the extracted features which are then trained using Split and Merge Expectation Maximization Algorithm that improves the global convergence of EM algorithm. It improves the learning of mixture models which in turn gives better LID performance. A maximum likelihood classifier is used for classification or identifying a language. The superiority of the proposed method is tested for four languages


Expectation Maximization Gaussian Mixture Model Expectation Maximization Algorithm Indian Language Gaussian Probability Density Function 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Muthusamy, Y.K.: A Segmental Approach to Automatic Language Identification. PhD thesis, Oregon Graduate Institute (1993)Google Scholar
  2. 2.
    Zissman, M.A., Singer, E.: Automatic language identification of telephone speech messages using phoneme recognition and N-GRAM modeling. In: Proc. ICASSP 1994, Adelaide, Austrailia (1994)Google Scholar
  3. 3.
    Nagrajan, T., Murthy, H.A.: Language identification using parallel syllable like unit recognition. In: Proc. ICASSP (2004)Google Scholar
  4. 4.
    Mary, L., Yegnanarayana, B.: Autoassociative Neural Network Models for Language ldentification. In: Proc. IClSlP (2004)Google Scholar
  5. 5.
    Hegde, R.M., Murthy, H.A.: Automatic Language Identification and Discrimination Using the Modified Group Delay Feature. In: Proc. ICISIP (2005)Google Scholar
  6. 6.
    Bimbot, F.E., Magrin-chagnolleau, I., Dutat, M.: Language recognition using time-frequency principal component analysis and acoustic modeling (2000)Google Scholar
  7. 7.
    Bilmes, J.A.: A gentle tutorial of the EM algorithm and its application to parameter estimation for gaussian mixture and hidden markov models. Technical Report tr-97-021, International Computer Science Institute, Berkeley, California, USA (1997)Google Scholar
  8. 8.
    Ueda, N., Nakano, R., Ghabramani, Z., Hinton, G.E.: SMEM algorithm for mixture models. Neural Computation (2000)Google Scholar
  9. 9.
    Cheng, S.S., Wang, H.M., Fu, H.C.: A model-selection-based self-splitting gaussian mixture learning with application to speaker identification. In: EURASIP (2004)Google Scholar
  10. 10.
    Ormoneit, D., Tresp, V.: Improved gaussian mixture density estimates using bayesian penalty terms and network averaging. In: NIPS (1995)Google Scholar
  11. 11.
    Zhang, Z., Chibiao Chen, J.S., Chan, K.L.: EM algorithms for gaussian mixtures with split-and-merge operation. Pattern Recognition (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Naresh Manwani
    • 1
  • Suman K. Mitra
    • 1
  • M. V. Joshi
    • 1
  1. 1.Dhirubhai Ambani Institute of Information and Communication Technology, GandhinagarIndia

Personalised recommendations