We propose "parse-filtering", a new approach to continuous speech recognition. With it, word sequence hypotheses generated on the basis of N-gram language models are verified by grammar-based parsing during the search for the best-scoring hypothesis, and unparsable hypotheses are filtered out immediately as the search proceeds. Experimental results show that this method yields a higher sentence accuracy than can be achieved with a trigram language model alone. Error reductions are, respectively, 10.0% for word error rate and 12.3% for sentence error rate.
Cite as: Hanazawa, K., Sakai, S. (2000) Continuous speech recognition with parse filtering. Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000), vol. 1, 262-265, doi: 10.21437/ICSLP.2000-65
@inproceedings{hanazawa00_icslp,
author={Ken Hanazawa and Shinsuke Sakai},
title={{Continuous speech recognition with parse filtering}},
year=2000,
booktitle={Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000)},
pages={vol. 1, 262-265},
doi={10.21437/ICSLP.2000-65}
}