In the conventional hidden Markov model, the model parameters are reestimated by an iterative procedure known as the Baum-Welch method. This paper proposes an alternative procedure using fuzzy estimation, which is generalised from the fuzzy c-means and the Baum-Welch methods. An extension of this approach, which uses a garbage state to deal with outlier data is also proposed. Experiments using the TI46 speech data corpus show this approach can be applicable to speech and speaker recognition.
Cite as: Tran, D., Wagner, M. (1999) Hidden Markov models using fuzzy estimation. Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999), 2749-2752, doi: 10.21437/Eurospeech.1999-605
@inproceedings{tran99_eurospeech,
author={Dat Tran and Michael Wagner},
title={{Hidden Markov models using fuzzy estimation}},
year=1999,
booktitle={Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999)},
pages={2749--2752},
doi={10.21437/Eurospeech.1999-605}
}