中国邮电高校学报(英文) ›› 2012, Vol. 19 ›› Issue (1): 94-100.doi: 10.1016/S1005-8885(11)60233-1

• Others • 上一篇    下一篇

Automatic context induction for tone model integration in mandarin speech recognition

黄浩,李兵虎   

  1. 新疆大学信息科学与工程学院
  • 收稿日期:2011-03-01 修回日期:2011-10-14 出版日期:2012-02-28 发布日期:2012-02-21
  • 通讯作者: 黄浩 E-mail:hwanghao@gmail.com
  • 基金资助:

    This work was supported by the National Natural Science Foundation of China (60965002), the College Research Project of Xinjiang (XJEDU2008S15), and the Start-up Fund Research for Ph. D. in Xinjiang University (BS090143).

Automatic context induction for tone model integration in mandarin speech recognition

  • Received:2011-03-01 Revised:2011-10-14 Online:2012-02-28 Published:2012-02-21
  • Contact: HUANG Hao E-mail:hwanghao@gmail.com
  • Supported by:

    This work was supported by the National Natural Science Foundation of China (60965002), the College Research Project of Xinjiang (XJEDU2008S15), and the Start-up Fund Research for Ph. D. in Xinjiang University (BS090143).

摘要:

Tone model (TM) integration is an important task for mandarin speech recognition. It has been proved to be effective to use discriminatively trained scaling factors when integrating TM scores into multi-pass speech recognition. Moreover, context-dependent (CD) scaling can be applied for better interpolation between the models. One limitation of this approach is a large number of parameters will be introduced, which makes the technique prone to overtraining. In this paper, we propose to induce context-dependent model weights by using automatically derived phonetic decision trees. Question at each tree node is chosen to minimize the expected recognition error on the training data. First order approximation of the minimum phone error (MPE) objective function is used for question pruning to make tree building efficient. Experimental results on continuous mandarin speech recognition show the method is capable of inducing the most crucial phonetic contexts and obtains significant error reduction with far fewer parameters, compared with that obtained by using manually designed context-dependent scaling parameters.

关键词:

TM integration, MPE, decision tree, mandarin speech recognition, context-dependent

Abstract:

Tone model (TM) integration is an important task for mandarin speech recognition. It has been proved to be effective to use discriminatively trained scaling factors when integrating TM scores into multi-pass speech recognition. Moreover, context-dependent (CD) scaling can be applied for better interpolation between the models. One limitation of this approach is a large number of parameters will be introduced, which makes the technique prone to overtraining. In this paper, we propose to induce context-dependent model weights by using automatically derived phonetic decision trees. Question at each tree node is chosen to minimize the expected recognition error on the training data. First order approximation of the minimum phone error (MPE) objective function is used for question pruning to make tree building efficient. Experimental results on continuous mandarin speech recognition show the method is capable of inducing the most crucial phonetic contexts and obtains significant error reduction with far fewer parameters, compared with that obtained by using manually designed context-dependent scaling parameters.

Key words:

TM integration, MPE, decision tree, mandarin speech recognition, context-dependent

中图分类号: