Graduate School and Research Center in Digital Sciences

Letter-to-sound pronunciation prediction using conditional random fields

Wang, Dong; King, Simon

Signal Processing Letters, Vol 18, N°2, February 2011

Pronunciation prediction, or letter-to-sound (LTS) conversion, is an essential task for speech synthesis, open vocabulary spoken term detection and other applications dealing with novel words. Most current approaches (at least for English) employ data-driven methods to learn and represent pronunciation rules using statistical models such as decision trees, hidden Markov models (HMMs) or joint-multigram models (JMMs). The LTS task remains challenging, particularly for languages with a complex relationship between spelling and pronunciation such as English. In this paper, we propose to use a conditional random field (CRF) to perform LTS because it avoids having to model a distribution over observations and can perform global inference, suggesting that it may be more suitable for LTS than decision trees, HMMs or JMMs. One challenge in applying CRFs to LTS is that the phoneme and grapheme sequences of a word are generally of different lengths, which makes CRF training difficult. To solve this problem, we employed a joint-multigram model to generate aligned training exemplars. Experiments conducted with the AMI05 dictionary demonstrate that a CRF significantly outperforms other models, especially if n-best lists of predictions are generated.

Document Doi Bibtex

Title:Letter-to-sound pronunciation prediction using conditional random fields
Keywords:Data models , Decision trees , Hidden Markov models , Joints , Markov processes , Predictive models , Training , conditional random field , joint multigram model , letter-to-sound , speech synthesis , spoken term detection
Type:Journal
Language:English
City:
Date:
Department:Data Science
Eurecom ref:3303
Copyright: © 2010 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Bibtex: @article{EURECOM+3303, doi = {http://dx.doi.org/10.1109/LSP.2010.2098440 }, year = {2010}, month = {12}, title = {{L}etter-to-sound pronunciation prediction using conditional random fields}, author = {{W}ang, {D}ong and {K}ing, {S}imon}, journal = {{S}ignal {P}rocessing {L}etters, {V}ol 18, {N}°2, {F}ebruary 2011}, url = {http://www.eurecom.fr/publication/3303} }