Training data selection for improving discriminative training of acoustic models

This paper considers training data selection for discriminative training of acoustic models for large vocabulary continuous speech recognition (LVCSR). Three novel data selection approaches are proposed. First, the average phone accuracy over all hypothesized word sequences in the word lattice of a...

Full description

Saved in:
Bibliographic Details
Published inPattern recognition letters Vol. 30; no. 13; pp. 1228 - 1235
Main Authors Chen, Berlin, Liu, Shih-Hung, Chu, Fang-Hui
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 01.10.2009
Elsevier
Subjects
Acoustic models
Applied sciences
Continuous speech recognition
Data selection
Discriminative training
Entropy
Exact sciences and technology
Information, signal and communications theory
Phone accuracy
Signal processing
Speech processing
Telecommunications and information theory
Acoustic models
Data selection
Entropy
Continuous speech recognition
Discriminative training
Phone accuracy
Performance evaluation
Discriminant analysis
Automatic transcription
Cable television
Learning
Accuracy
Posterior probability
Audiovisual document
News
Speech recognition
Speech processing
Online AccessGet full text

Cover

More Information
Summary:This paper considers training data selection for discriminative training of acoustic models for large vocabulary continuous speech recognition (LVCSR). Three novel data selection approaches are proposed. First, the average phone accuracy over all hypothesized word sequences in the word lattice of a training utterance is utilized for utterance-level data selection. Second, phone-level data selection based on the difference between the expected accuracy of a phone arc and the average phone accuracy of the word lattice is investigated. Finally, frame-level data selection based on the normalized frame-level entropy of Gaussian posterior probabilities obtained from the word lattice is explored. The underlying characteristics of the presented approaches are extensively investigated and their performance is verified by comparison with standard discriminative training approaches. Experiments conducted on a broadcast news speech transcription task show that with the aid of phone- and frame-level data selection we can reduce more than half of the turnaround time for acoustic model training and simultaneously obtain a comparably good set of discriminative acoustic models.
ISSN:0167-8655
1872-7344
DOI:10.1016/j.patrec.2009.05.009