Speaker adaptation by hierarchical EigenVoice

@inproceedings{Onishi2003SpeakerAB,
  title={Speaker adaptation by hierarchical EigenVoice},
  author={Yoshifumi Onishi and Ken-ichi Iso},
  booktitle={ICASSP},
  year={2003}
}
We propose a novel speaker adaptation method, Hierarchical EigenVoice (HEV). This method extends the EigenVoice [ I ] through clustering the Gaussian components of HMMs into a hierarchical tree structure. It enables to autonomously control a number of adaptation parameters (model complexity) depending on the amount of adaptation utterances from a new speaker. The experimental results of Japanese large vocabulary continuous speech recognition confirmed the significant performance increase in all… CONTINUE READING

From This Paper

Figures, tables, and topics from this paper.
5 Citations
7 References
Similar Papers

References

Publications referenced by this paper.
Showing 1-7 of 7 references

Bottenveck “ Anisotropic map defined by eigenvoices for large vocabulary continuous speech recognition

  • H.
  • Proc . Inf . Conf Acousrics , Speech , Signal…
  • 2001

Niedzielski “ Rapid Speaker Adaptation in Eigenvoice Space

  • N.
  • IEEE Trans . Speech Audio Proc .
  • 1994

Watanabe “ Speaker adaptation with autonomous control using tree structure

  • T.

Similar Papers

Loading similar papers…