Conference paper
Discriminative training of self-structuring hidden control neural models
This paper presents a new training algorithm for self-structuring hidden control neural (SHC) models. The SHC models were trained non-discriminatively for speech recognition applications. Better recognition performance can generally be achieved, if discriminative training is applied instead. Thus we developed a discriminative training algorithm for SHC models, where each SHC model for a specific speech pattern is trained with utterances of the pattern to be recognized and with other utterances.
The discriminative training of SHC neural models has been tested on the TIDIGITS database
Language: | English |
---|---|
Publisher: | IEEE |
Year: | 1995 |
Pages: | 3379-3382 |
Proceedings: | 1995 IEEE International Conference on Acoustics, Speech, and Signal Processing |
ISBN: | 0780324315 and 9780780324312 |
ISSN: | 2379190x and 15206149 |
Types: | Conference paper |
DOI: | 10.1109/ICASSP.1995.479710 |
ORCIDs: | Sørensen, Helge Bjarup Dissing |
Databases Equations Hidden Markov models Neural networks Neurons Pattern recognition Predictive models SHC neural models Speech recognition TIDIGITS database Testing Vocabulary discriminative training algorithm hidden Markov models learning (artificial intelligence) neural nets recognition performance self-structuring hidden control neural models speech pattern speech recognition speech recognition applications