Pinched Lattice Minimum Bayes Risk Discriminative Training for Large Vocabulary Continuous Speech Recognition

Download: PDF, slides.

“Pinched Lattice Minimum Bayes Risk Discriminative Training for Large Vocabulary Continuous Speech Recognition” by V. Doumpiotis and W. Byrne. In Proc. of the International Conference on Spoken Language Processing, 2004.

Abstract

Iterative estimation procedures that minimize empirical risk based on general loss functions such as the Levenshtein distance have been derived as extensions of the Extended Baum Welch algorithm. While reducing expected loss on training data is a desirable training criterion, these algorithms can be difficult to apply. They are unlike MMI estimation in that they require an explicit listing of the hypotheses to be considered and in complex problems such lists tend to be prohibitively large. To overcome this difficulty, modeling techniques originally developed to improve search efficiency in Minimum Bayes Risk decoding can be used to transform these estimation algorithms so that exact update, risk minimization procedures can be used for complex recognition problems. Experimental results in two large vocabulary speech recognition tasks show improvements over conventionally trained MMIE models.

Download: PDF, slides.

BibTeX entry:

@inproceedings{icslp04_plmbrdt,
   author = {V. Doumpiotis and W. Byrne},
   title = {Pinched Lattice Minimum {B}ayes Risk Discriminative Training
	for Large Vocabulary Continuous Speech Recognition},
   booktitle = {Proc. of the International Conference on Spoken Language
	Processing},
   pages = {(4 pages)},
   year = {2004}
}

Back to Bill Byrne publications.