Human listeners use lexical stress for word segmentation and disambiguation. We look into using lexical stress for large-vocabulary speech recognition for the Dutch language. It appears that beside vowels, consonants should be taken into account. By introducing stressed phonemes, and features for spectral bands and the fundamental frequency, we reduce the word error rate by 2.6%.
Cite as: Dalen, R.C.v., Wiggers, P., Rothkrantz, L.J.M. (2006) Lexical stress in continuous speech recognition. Proc. Interspeech 2006, paper 1578-Thu1FoP.7, doi: 10.21437/Interspeech.2006-597
@inproceedings{dalen06_interspeech,
author={Rogier C. van Dalen and Pascal Wiggers and Léon J. M. Rothkrantz},
title={{Lexical stress in continuous speech recognition}},
year=2006,
booktitle={Proc. Interspeech 2006},
pages={paper 1578-Thu1FoP.7},
doi={10.21437/Interspeech.2006-597}
}