We present an online discriminative training approach to graphemeto- phoneme (g2p) conversion. We employ a many-to-many alignment between graphemes and phonemes, which overcomes the limitations of widely used one-to-one alignments. The discriminative structure-prediction model incorporates input segmentation, phoneme prediction, and sequence modeling in a unified dynamic programming framework. The learning model is able to capture both local context features in inputs, as well as non-local dependency features in sequence outputs. Experimental results show that our system surpasses the state-of-the-art on several data sets.
Cite as: Jiampojamarn, S., Kondrak, G. (2009) Online discriminative training for grapheme-to-phoneme conversion. Proc. Interspeech 2009, 1303-1306, doi: 10.21437/Interspeech.2009-407
@inproceedings{jiampojamarn09_interspeech,
author={Sittichai Jiampojamarn and Grzegorz Kondrak},
title={{Online discriminative training for grapheme-to-phoneme conversion}},
year=2009,
booktitle={Proc. Interspeech 2009},
pages={1303--1306},
doi={10.21437/Interspeech.2009-407}
}