An automatic speaker identification system is developed based on a neural predictive HMM speech recognition system. The neural predictive HMM utilises layered feed-forward neural networks to implement joint linear/nonlinear speech-frame prediction. A Markov chain is used to control changes in the network's weight parameters. The system exhibits high accuracy speaker recognition using very short speech utterances. The system was also shown to be robust against intraspeaker speech variations. The developed system is also well suited for efficient parallel V1SI implementations enabling real time speaker identification performance.
Cite as: Hassanein, K., Deng, L., Elmasry, M.I. (1994) A neural predictive hidden Markov model for speaker recognition. Proc. ESCA Workshop on Automatic Speaker Recognition, Identification and Verification, 115-118
@inproceedings{hassanein94_asriv,
author={K. Hassanein and L. Deng and M. I. Elmasry},
title={{A neural predictive hidden Markov model for speaker recognition}},
year=1994,
booktitle={Proc. ESCA Workshop on Automatic Speaker Recognition, Identification and Verification},
pages={115--118}
}