A new hybrid decoding algorithm for speech recognition and utterance verification

We propose a new search algorithm for one-pass utterance verification in which verification and recognition are simultaneously done. The basic idea is that we make a hybrid decoder in which the conventional Viterbi decoder is combined with a likelihood ratio decoder based on a confidence score repre...

Full description

Saved in:
Bibliographic Details
Main Authors: Myoung-Wan Koo, Chin-Hui Lee, Biing-Hwang Juang
Format: Conference Proceeding
Language:English
Subjects:
Decoding
Dynamic range
Error correction
Multimedia communication
Robustness
Signal design
Speech recognition
Telecommunications
Testing
Viterbi algorithm
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We propose a new search algorithm for one-pass utterance verification in which verification and recognition are simultaneously done. The basic idea is that we make a hybrid decoder in which the conventional Viterbi decoder is combined with a likelihood ratio decoder based on a confidence score representing the confidence level for utterance verification. We overcome the dynamic range problem encountered in the likelihood ratio decoder by use of a sigmoid limiter. We also improve its robustness by selective use of the anti-phone models in calculating the phone likelihood ratios. The proposed strategy is applied to a 1000-word car reservation task. Experimental results show that the proposed hybrid decoder gives better results than the conventional likelihood decoder or the likelihood ratio decoder, particularly in dealing with out-of-vocabulary words or out-of-task sentences.
DOI:10.1109/ASRU.1997.659104