Title :
A decision theorectic formulation of a training problem in speech recognition and a comparison of training by unconditional versus conditional maximum likelihood
Author_Institution :
IBM T.J. Watson Research Center, Yorktown Heights, NY
fDate :
8/1/1983 12:00:00 AM
Abstract :
The choice of method for training a speech recognizer is posed as an optimization problem. The currently used method of maximum likelihood, while heuristic, is shown to be superior under certain assumptions to another heuristic: the method of conditional maximum likelihood.
Keywords :
Automatic speech recognition; Feature extraction; Maximum likelihood decoding; Microphones; Optimization methods; Predictive models; Signal processing; Speech recognition; Statistical analysis; Vectors;
Journal_Title :
Acoustics, Speech and Signal Processing, IEEE Transactions on
DOI :
10.1109/TASSP.1983.1164173
