ISCA Archive Interspeech 2008
ISCA Archive Interspeech 2008

Minimum phone error discriminative training for Mandarin Chinese speaker adaptation

Liang-Yu Chen, Chun-Jen Lee, Jyh-Shing Roger Jang

Speaker adaptation is an efficient way to model a new speaker from an existing speaker-independent model with limited speakerdependent data. In this paper, we investigate the use of discriminative training schemes based on the minimum phone error (MPE) criterion to improve a well-known speaker adaptation technique, a combination of transform-based adaptation and Bayesian adaptation. Furthermore, a new approach utilizing the statistics of the model-based regression tree for controlling the interpolation between maximum likelihood (ML) and MPE objective functions is also presented. Several comparative experiments were conducted on a continuous speech recognition task for Mandarin Chinese. Experimental results show that the proposed approach can further improve the performance of the original hybrid adaptation.


doi: 10.21437/Interspeech.2008-376

Cite as: Chen, L.-Y., Lee, C.-J., Jang, J.-S.R. (2008) Minimum phone error discriminative training for Mandarin Chinese speaker adaptation. Proc. Interspeech 2008, 1241-1244, doi: 10.21437/Interspeech.2008-376

@inproceedings{chen08e_interspeech,
  author={Liang-Yu Chen and Chun-Jen Lee and Jyh-Shing Roger Jang},
  title={{Minimum phone error discriminative training for Mandarin Chinese speaker adaptation}},
  year=2008,
  booktitle={Proc. Interspeech 2008},
  pages={1241--1244},
  doi={10.21437/Interspeech.2008-376}
}