Generalized BaumWelch Algorithm for Discriminative Training on Large Vocabulary Continuous Speech Recognition System (2009)
| Venue: | in Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing |
| Citations: | 4 - 4 self |
BibTeX
@INPROCEEDINGS{Hsiao09generalizedbaumwelch,
author = {Roger Hsiao and Yik-cheung Tam and Tanja Schultz},
title = {Generalized BaumWelch Algorithm for Discriminative Training on Large Vocabulary Continuous Speech Recognition System},
booktitle = {in Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing},
year = {2009}
}
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
We propose a new optimization algorithm called Generalized Baum Welch (GBW) algorithm for discriminative training on hidden Markov model (HMM). GBW is based on Lagrange relaxation on a transformed optimization problem. We show that both Baum-Welch (BW) algorithm for ML estimate of HMM parameters, and the popular extended Baum-Welch (EBW) algorithm for discriminative training are special cases of GBW. We compare the performance of GBW and EBW for Farsi large vocabulary continuous speech recognition (LVCSR). Index Terms — Speech recognition, discriminative training.
Citations
| 2134 | Convex Optimization - Boyd, Vandenberghe - 2004 |
| 71 | Discriminative Training for Large Vocabulary Speech Recognition - Povey - 2004 |
| 36 | An improved mmie training algorithm for speaker-independent, small vocabulary, continuous speech recognition - Normandin, Morgera - 1991 |
| 33 | Large margin hidden markov models for automatic speech recognition,” Advances in neural information processing systems - Sha, Saul |
| 32 | MMIE training of large vocabulary recognition systems - Valtchev, Odell, et al. - 1997 |
