A new parameter smoothing method in the hybrid TDNN/HMM architecture for speech recognition.

C. S. Jang C. K. Un

Published in: Speech Commun. (1996)

Keyphrases

speech recognition
hidden markov models
pattern recognition
mobile devices
edge detection
language model
speech recognizer
speech recognition technology
support vector machine
maximum likelihood
hybrid method
automatic speech recognition
speaker recognition
speech synthesis
speaker adaptation
speech recognizers