A more efficient and optimal LLR for decoding and verification

Author

Leung, Lam Kwok ; Fung, Pascale

Author_Institution

Human Language Technol. Center, Univ. of Sci. & Technol., Clear Water Bay, Hong Kong

Volume

2

fYear

1999

fDate

15-19 Mar 1999

Firstpage

689

Abstract

We propose a new confidence score for decoding and verification. Since the traditional log likelihood ratio (LLR) is borrowed from speaker verification technique, it may not be appropriate for decoding because we do not have a good modelling and definition of LLR for decoding/utterance verification. We have proposed a new formulation of LLR that can be used for decoding and verification task. Experimental results show that our proposed LLR can perform equally well compared with the result based on maximum likelihood in a decoding task. Also, we get an 5% improvement in decoding compared with traditional LLR

Keywords

Viterbi decoding; optimisation; speaker recognition; confidence score; decoding; efficient LLR; experimental results; keyword spotting; log likelihood ratio; maximum likelihood; modified Viterbi decoder; optimal LLR; speaker verification; speech recognition; spontaneous speech; utterance verification; Degradation; Hidden Markov models; Humans; Maximum likelihood decoding; Maximum likelihood detection; Speech processing; Speech recognition; Telephony; Viterbi algorithm;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1999. Proceedings., 1999 IEEE International Conference on

Conference_Location

Phoenix, AZ

ISSN

1520-6149

Print_ISBN

0-7803-5041-3

Type

conf

DOI

10.1109/ICASSP.1999.759760

Filename

759760