This paper describes a robust speech detection algorithm for speech-activated hands-free applications. The system consists of three techniques: (1) noise suppression with efficient implementation, (2) robust endpoint detection and (3) speech verification using garbage modeling and confidence measure. With efficient implementation, noise suppression improves the SNR by roughly 10-20 dB. The endpoint detection uses the technique described in [1] with improvement for non-stationary noise. Garbage modeling and confidence measure are used to handle out-of-vocabulary (OOV) words and background pulse noise.
Wu, D., M. Tanaka, R. Chen and L. Olorenshaw, "A Robust Endpoint Detection Algorithm for Speech Recognition in Cars" Proceedings-97 of Sony Research Forum, Tokyo, 1997.
Cite as: Wu, D., Menendez-Pidal, X., Olorenshaw, L., Chen, R., Tanaka, M., Amador, M. (2000) Speech and word detection algorithms for hands-free applications. Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000), vol. 4, 398-401, doi: 10.21437/ICSLP.2000-834
@inproceedings{wu00i_icslp,
author={Duanpei Wu and X. Menendez-Pidal and L. Olorenshaw and R. Chen and M. Tanaka and M. Amador},
title={{Speech and word detection algorithms for hands-free applications}},
year=2000,
booktitle={Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000)},
pages={vol. 4, 398-401},
doi={10.21437/ICSLP.2000-834}
}