Minimum Bayes-Risk Automatic Speech Recognition

Minimum Bayes-Risk Automatic Speech Recognition” by V. Goel and W. Byrne. Computer Speech and Language, vol. 14(2), 2000, pp. 115-135 (21 pages).

Abstract

In this paper we address the problem of efficient implementation of the minimum Bayes-risk classifiers for automatic speech recognition. Simplifying assumptions that allow computationally feasible approximations to these classifiers are proposed. Under these assumptions an approximate implementation as an A-star search algorithm over recognition lattice is constructed. This algorithm improves up on the previously proposed N-best list rescoring implementation of these classifiers. The minimum Bayes-risk classifiers are shown to outperform the most commonly used maximum a-posteriori probability (MAP) classifier on three speech recognition tasks: reduction of word error rate, reduction of content word error rate, and identification of Named Entities in speech. The A-star implementation is also contrasted with the N-best list rescoring implementation and is found to obtain modest but significant improvements in accuracy with little computational overhead.

BibTeX entry:

@article{mbr_csl00,
   author = {V. Goel and W. Byrne},
   title = {Minimum {B}ayes-{R}isk Automatic Speech Recognition},
   journal = {Computer Speech and Language},
   volume = {14(2)},
   pages = {115--135 (21 pages)},
   year = {2000},
   url = {http://dx.doi.org/10.1006/csla.2000.0138}
}

Back to Bill Byrne publications.