This paper addresses the problem of mismatch between training and testing conditions in a HMM-based speech recognizer. Parallel Model Combination (PMC) has demonstrated to be an efficient technique for reducing the effects of additive noise. In order to apply this technique, a noise HMM must be trained at the recognition phase. Approaches that estimate the noise model based on the Expectation-Maximization (EM) or Baum-Welch algorithms are widely used. In these methods the recorded environmental noise data are used, and their major drawback is that they need a long sequence of noise data to estimate properly the model parameters. In some real life applications the amount of noise data can be too small, so from a practical point of view, the needed amount of noise is a critical parameter which should be as short as possible. We propose a novel method for obtaining a more reliable noise model than training it from scratch by using a short noise sequence.
Cite as: Docio-Fernández, L., García-Mateo, C. (1998) Noise model selection for robust speech recognition. Proc. 5th International Conference on Spoken Language Processing (ICSLP 1998), paper 0579, doi: 10.21437/ICSLP.1998-321
@inproceedings{dociofernandez98_icslp,
author={Laura Docio-Fernández and Carmen García-Mateo},
title={{Noise model selection for robust speech recognition}},
year=1998,
booktitle={Proc. 5th International Conference on Spoken Language Processing (ICSLP 1998)},
pages={paper 0579},
doi={10.21437/ICSLP.1998-321}
}