Advertisement

Recurrent Neural Network Based Phoneme Recognition Incorporating Articulatory Dynamic Parameters

  • Mohammed Rokibul Alam Kotwal
  • Foyzul Hassan
  • Md. Mahabubul Alam
  • Abdur Rahman Khan Jehad
  • Md. Arifuzzaman
  • Mohammad Nurul Huda
Part of the Communications in Computer and Information Science book series (CCIS, volume 192)

Abstract

This paper describes a recurrent neural network (RNN) based phoneme recognition method incorporating articulatory dynamic parameters (Δ and ΔΔ). The method comprises three stages: (i) DPFs extraction using a recurrent neural network (RNN) from acoustic features, (ii) incorporation of dynamic parameters into a multilayer neural network (MLN) for reducing DPF context, and (iii) addition of an Inhibition/Enhancement (In/En) network for categorizing the DPF movement more accurately and Gram-Schmidt orthogonalization procedure for decorrelating the inhibited/enhanced data vector before connecting with a hidden Markov models (HMMs)-based classifier. From the experiments on Japanese Newspaper Article Sentences (JNAS), it is observed that the proposed method provides a higher phoneme correct rate over the method that does not incorporate dynamic articulatory parameters. Moreover, it reduces mixture components in HMM for obtaining a higher performance.

Keywords

Distinctive Phonetic Feature Multi-Layer Neural Network Recurrent Neural Network Inhibition/Enhancement Network Local Features 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Kirchhoff, K., et al.: Combining acoustic and articulatory feature information for robust speech recognition. Speech Commun. 37, 303–319 (2002)CrossRefzbMATHGoogle Scholar
  2. 2.
    Kirchhoffs, K.: Robust Speech Recognition Using Articulatory information. Ph.D thesis, University of Bielefeld, Germany (July 1999)Google Scholar
  3. 3.
    Leung, K.Y., Mak, M.W., Kung, S.Y.: Applying articulatory features to telephone-based speaker verification. In: Proc. IEEE ICASSP 2004, vol. I, pp. 85–88 (2004)Google Scholar
  4. 4.
    Fukuda, T., Yamamoto, W., Nitta, T.: Distinctive Phonetic feature Extraction for robust speech recognition. In: Proc. ICASSP 2003, vol. II, pp. 25–28 (2003)Google Scholar
  5. 5.
    Fukuda, T., Nitta, T.: Orthogonalized Distinctive Phonetic feature Extraction for Noise-Robust Automatic Speech Recognition. The Institute of Electronics, Information and Communication Engineers (IEICE) Transactions on Information and Systems E87-D(5), 1110–1118 (2004)Google Scholar
  6. 6.
    Huda, et al.: Distinctive Phonetic Feature (DPF) based phone segmentation using 2-stage multilayer neural network. In: NCSP 2007, Shanghai, China ( March 2007)Google Scholar
  7. 7.
    Ansary, L., et al.: Modeling phones coarticulation effects in a neural network based speech recognition system. In: Proc. Interspeech (2004)Google Scholar
  8. 8.
    Robinson, T.: An application of Recurrent Nets to Phone Probability Estimation. IEEE Trans. Neural Networks  5(3) (1994)Google Scholar
  9. 9.
    Huda, M.N., et al.: Phoneme recognition based on hybrid neural network with inhibition/enhancement of distinctive phonetic feature (DPF) trajectories. In: InterSpeech 2008, Brisbane, Australia (September 2008)Google Scholar
  10. 10.
    King, S., Taylor, P.: Detection of Phonological Features in Continuous Speech using Neural Networks. Computer Speech and Language 14(4), 333–345 (2000)CrossRefGoogle Scholar
  11. 11.
    Eide, E.: Distinctive Features for Use in an Automatic Speech Recognition System. In: Proc. Eurospeech 2001, vol. III, pp. 1613–1616 (2001)Google Scholar
  12. 12.
    Nitta, T.: Feature extraction for speech recognition based on orthogonal acoustic-feature planes and LDA. In: Proc. ICASSP 1999, pp. 421–424 (1999)Google Scholar
  13. 13.
    Kobayashi, T., et al.: ASJ Continuous Speech Corpus for Research. Acoustic Society of Japan Trans. 48(12), 888–893 (1992)Google Scholar
  14. 14.
    JNAS: Japanese Newspaper Article Sentences, http://www.milab.is.tsukuba.ac.jp/jnas/instruct.html

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Mohammed Rokibul Alam Kotwal
    • 1
  • Foyzul Hassan
    • 1
  • Md. Mahabubul Alam
    • 1
  • Abdur Rahman Khan Jehad
    • 2
  • Md. Arifuzzaman
    • 2
  • Mohammad Nurul Huda
    • 1
  1. 1.United International UniversityDhakaBangladesh
  2. 2.The University of Asia PacificDhakaBangladesh

Personalised recommendations