Login / Signup
A new training method for multi-phone speech units for use in a hidden Markov model speech recognition system.
Jade Goldstein
Akio Amano
Hideki Murayama
Mariko Izawa
Akira Ichikawa
Published in:
ICSLP (1990)
Keyphrases
</>
hidden markov models
speech recognition
markov model
dynamic programming
speaker recognition
discriminative training
conditional random fields
speech signal
hybrid method
forward backward
mixture distributions