Publication Details
Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments
Zweig Geoffrey
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Schwarz Petr, Ing., Ph.D. (DCGM)
Heřmanský Hynek, prof. Ing., Dr. Eng. (DCGM)
speech recognition
the paper is on confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments
Automatic Speech Recognition (ASR) systems continue to make errors during search when handling various phenomena including noise, pronunciation variation, and out of vocabulary (OOV) words. Predicting the probability that a word is incorrect can prevent the error from propagating and perhaps allow the system to recover. This paper addresses the problem of detecting errors and OOVs for read Wall Street Journal speech when the word error rate (WER) is very low. It augments a traditional confidence estimate by introducing two novel methods: phone-level comparison using Multi-String Alignment (MSA) and word-level comparison using phone-to-word transduction. We show that features from phone and word string comparisons can be added to a standard maximum entropy framework thereby substantially improving performance in detecting both errors and OOVs. Additionally we show an extension to detecting English and accented English for the Language Identification (LID) task.
@inproceedings{BUT32069,
author="Christopher {White} and Geoffrey {Zweig} and Lukáš {Burget} and Petr {Schwarz} and Hynek {Heřmanský}",
title="Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments",
booktitle="Proc. 2008 IEEE International Conference on Acoustics, Speech, and Signal Processing",
year="2008",
pages="1--4",
publisher="IEEE Signal Processing Society",
address="Las Vegas",
isbn="1-4244-1484-9",
url="http://www.fit.vutbr.cz/research/groups/speech/publi/2008/white_icassp08.pdf"
}