To achieve high recognition performance for a wide variety of noise and for a wide range of signal-to-noise ratio, this paper presents integration methods of four noise reduction algorithms: spectral subtraction with smoothing of time direction, temporal domain SVD-based speech enhancement, GMM-based speech estimation and KLT-based comb-filtering. In this paper, we proposed two types of ?combination methods of noise suppression algorithms: selection of front-end processor and combination of results from multiple recognition processes. Recognition results on the AURORA-2J task showed the effectiveness of our proposed methods.
Cite as: Kitaoka, N., Hamaguchi, S., Nakagawa, S. (2006) Noisy speech recognition based on selection of multiple noise suppression methods using noise GMMs. Proc. Interspeech 2006, paper 1207-Thu2CaP.12, doi: 10.21437/Interspeech.2006-643
@inproceedings{kitaoka06_interspeech,
author={Norihide Kitaoka and Souta Hamaguchi and Seiichi Nakagawa},
title={{Noisy speech recognition based on selection of multiple noise suppression methods using noise GMMs}},
year=2006,
booktitle={Proc. Interspeech 2006},
pages={paper 1207-Thu2CaP.12},
doi={10.21437/Interspeech.2006-643}
}