ICASSP
2010
IEEE
13 years 2 months ago
2010
IEEE
We describe a fast and efficient online algorithm for phoneme sequence speech recognition. Our method is using a discriminative training to update the model parameters one utteran...