Advertisement

LSTM Deep Neural Networks Postfiltering for Improving the Quality of Synthetic Voices

  • Marvin Coto-JiménezEmail author
  • John Goddard-Close
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9703)

Abstract

Recent developments in speech synthesis have produced systems capable of providing intelligible speech, and researchers now strive to create models that more accurately mimic human voices. One such development is the incorporation of multiple linguistic styles in various languages and accents. HMM-based speech synthesis is of great interest to researchers, due to its ability to produce sophisticated features with a small footprint. Despite such progress, its quality has not yet reached the level of the current predominant unit-selection approaches, that select and concatenate recordings of real speech. Recent efforts have been made in the direction of improving HMM-based systems. In this paper, we present the application of long short-term memory deep neural networks as a postfiltering step in HMM-based speech synthesis. Our motivation stems from a desire to obtain spectral characteristics closer to those of natural speech. The results described in the paper indicate that HMM-voices can be improved using this approach.

Keywords

LSTM HMM Speech synthesis Statistical parametric speech synthesis Postfiltering Deep learning 

Notes

Acknowledgements

This work was supported by the SEP and CONACyT under the Program SEP-CONACyT, CB-2012-01, No.182432, in Mexico, as well as the University of Costa Rica in Costa Rica.

References

  1. 1.
    Tokuda, K., Nankaku, Y., Toda, T., Zen, H., Yamagishi, J., Oura, K.: Speech synthesis based on hidden markov models. Proc. IEEE 101(5), 1234–1252 (2013)CrossRefGoogle Scholar
  2. 2.
    Black, A.W.: Unit selection and emotional speech. In: Interspeech (2003)Google Scholar
  3. 3.
    Yoshimura, T., Tokuda, T., Masuko, T., Kobayashi, T., Kitamura, T.: Simultaneous modeling of spectrum, pitch and duration in HMM-based speech synthesis. In: Proceedings of the Eurospeech, pp. 2347–2350 (1999)Google Scholar
  4. 4.
    Falaschi, A., Giustiniani, M., Verola, M.: A hidden markov model approach to speech synthesis. In: Proceedings of the Eurospeech, pp. 2187–2190 (1989)Google Scholar
  5. 5.
    Karabetsos, S., Tsiakoulis, P., Chalamandaris, A., Raptis, S.: HMM-based speech synthesis for the greek language. In: Sojka, P., Horák, A., Kopeček, I., Pala, K. (eds.) TSD 2008. LNCS (LNAI), vol. 5246, pp. 349–356. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  6. 6.
    Pucher, M., Schabus, D., Yamagishi, Y., Neubarth, F., Strom, V.: Modeling and interpolation of austrian german and viennese dialect in HMM-based speech synthesis. Speech Commun. 52(2), 164–179 (2010)CrossRefGoogle Scholar
  7. 7.
    Erro, D., Sainz, I., Luengo, I., Odriozola, I., Sánchez, J., Saratxaga, I., Navas, E., Hernáez, I.: HMM-based speech synthesis in basque language using HTS. In: Proceedings of the FALA (2010)Google Scholar
  8. 8.
    Stan, A., Yamagishi, Y., King, S., Aylett, M.: The romanian speech synthesis (RSS) corpus: building a high quality HMM-based speech synthesis system using a high sampling rate. Speech Commun. 53(3), 442–450 (2011)CrossRefGoogle Scholar
  9. 9.
    Kuczmarski, T.: HMM-based speech synthesis applied to polish. Speech Lang. Technol. 12, 13 (2010)Google Scholar
  10. 10.
    Hanzlíček, Z.: Czech HMM-based speech synthesis. In: Sojka, P., Horák, A., Kopeček, I., Pala, K. (eds.) TSD 2010. LNCS, vol. 6231, pp. 291–298. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  11. 11.
    Li, Y., Pan, S., Tao, J.: HMM-based speech synthesis with a flexible mandarin stress adaptation model. In: Proceedings of the 10th ICSP2010 Proceedings, Beijing, pp. 625–628 (2010)Google Scholar
  12. 12.
    Phan, S.T., Vu, T.T., Duong, C.T., Luong, M.C.: A study in vietnamese statistical parametric speech synthesis based on HMM. Int. J. 2(1), 1–6 (2013)MathSciNetGoogle Scholar
  13. 13.
    Boothalingam, R., Sherlin, S.V., Gladston, A.R., Christina, S.L., Vijayalakshmi, P., Thangavelu, N., Murthy, H.A.: Development and evaluation of unit selection and HMM-based speech synthesis systems for Tamil. In: National Conference on Communications (NCC), pp. 1–5. IEEE (2013)Google Scholar
  14. 14.
    Khalil, K.M., Adnan, C.: Implementation of speech synthesis based on HMM using PADAS database. In: 12th International Multi-Conference on Systems, Signals & Devices (SSD), pp. 1–6. IEEE (2015)Google Scholar
  15. 15.
    Nakamura, K., Oura, K., Nankaku, Y., Tokuda, K.: HMM-based singing voice synthesis and its application to japanese and english. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 265–269 (2014)Google Scholar
  16. 16.
    Roekhaut, S., Brognaux, S., Beaufort, R., Dutoit, T.: Elite-HTS: a NLP tool for French HMM-based speech synthesis. In: Interspeech, pp. 2136–2137 (2014)Google Scholar
  17. 17.
    HMM-based Speech Synthesis System (HTS). http://hts.sp.nitech.ac.jp/
  18. 18.
    Chen, L.H., Raitio, T., Valentini-Botinhao, C., Ling, Z.H., Yamagishi, J.: A deep generative architecture for postfiltering in statistical parametric speech synthesis. IEEE/ACM Trans. Audio, Speech Lang. Process. (TASLP) 23(11), 2003–2014 (2015)CrossRefGoogle Scholar
  19. 19.
    Takamichi, S., Toda, T., Neubig, G., Sakti, S., Nakamura, S.: A postfilter to modify the modulation spectrum in HMM-based speech synthesis. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 290–294 (2014)Google Scholar
  20. 20.
    Takamichi, S., Toda, T., Black, A.W., Nakamura, S.: Modified post-filter to recover modulation spectrum for HMM-based speech synthesis. In: IEEE Global Conference on Signal and Information Processing (GlobalSIP), pp. 547–551 (2014)Google Scholar
  21. 21.
    Prasanna, K.M., Black, A.W.: Recurrent Neural Network Postfilters for Statistical Parametric Speech Synthesis. arXiv preprint (2016). arXiv:1601.07215
  22. 22.
    Fan, Y., Qian, Y., Xie, F.L., Soong, F.K.: TTS synthesis with bidirectional LSTM based recurrent neural networks. In: Interspeech, pp. 1964–1968 (2014)Google Scholar
  23. 23.
    Zen, H., Sak, H.: Unidirectional long short-term memory recurrent neural network with recurrent output layer for low-latency speech synthesis. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 4470–4474 (2015)Google Scholar
  24. 24.
    Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997)CrossRefGoogle Scholar
  25. 25.
    Graves, A., Jaitly, N., Mohamed, A.: Hybrid speech recognition with deep bidirectional LSTM. In: IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) (2013)Google Scholar
  26. 26.
    Graves, A., Fernández, S., Schmidhuber, J.: Bidirectional LSTM networks for improved phoneme classification and recognition. In: Duch, W., Kacprzyk, J., Oja, E., Zadrożny, S. (eds.) ICANN 2005. LNCS, vol. 3697, pp. 799–804. Springer, Heidelberg (2005)Google Scholar
  27. 27.
    Erro, D., Sainz, I., Navas, E., Hernaez, I.: Improved HNM-based vocoder for statistical synthesizers. In: InterSpeech, pp. 1809–1812 (2011)Google Scholar
  28. 28.
    Kominek, J., Black, A.W.: The CMU Arctic speech databases. In: Fifth ISCA Workshop on Speech Synthesis (2004)Google Scholar
  29. 29.
    Zen, H., Senior, A., Schuster, M.: Statistical parametric speech synthesis using deep neural networks. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2013)Google Scholar
  30. 30.
    Zen, H., Senior, A.: Deep mixture density networks for acoustic modeling in statistical parametric speech synthesis. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2014)Google Scholar
  31. 31.
    Kominek, J., Schultz, T., Black, A.W.: Synthesizer voice quality of new languages calibrated with mean mel cepstral distortion. In: SLTU (2008)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.University of Costa RicaSan JoséCosta Rica
  2. 2.Autonomous Metropolitan UniversityMexicoMexico

Personalised recommendations