ISCA Archive Interspeech 2005
ISCA Archive Interspeech 2005

Background model based posterior probability for measuring confidence

Peng Liu, Ye Tian, Jian-Lai Zhou, Frank K. Soong

Word posterior probability (WPP) computed over LVCSR word graphs has been used successfully in measuring confidence of speech recognition output. However, for certain applications the word graph is too sparse to warrant reliable WPP estimation. In this paper, we incorporate subword units as background models to generate a subword graph for estimating posterior probability. Experiments on both English and Chinese databases show that syllable background models can repopulate the dynamic hypothesis space for effective computation of confidence measure. The resultant posterior probability confidence measure achieves 94.3% and 95.2% Out-Of-Vocabulary (OOV) word detection / rejection in English and Chinese, respectively. Correspondingly, confidence error rates are at 6.0% and 6.4%, respectively.

doi: 10.21437/Interspeech.2005-518

Cite as: Liu, P., Tian, Y., Zhou, J.-L., Soong, F.K. (2005) Background model based posterior probability for measuring confidence. Proc. Interspeech 2005, 1465-1468, doi: 10.21437/Interspeech.2005-518

  author={Peng Liu and Ye Tian and Jian-Lai Zhou and Frank K. Soong},
  title={{Background model based posterior probability for measuring confidence}},
  booktitle={Proc. Interspeech 2005},