To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used. For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination. We also investigate the robustness of matched and mismatched noise conditions between training and evaluation environments.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Makoto SAKAI, Norihide KITAOKA, Yuya HATTORI, Seiichi NAKAGAWA, Kazuya TAKEDA, "Evaluation of Combinational Use of Discriminant Analysis-Based Acoustic Feature Transformation and Discriminative Training" in IEICE TRANSACTIONS on Information,
vol. E93-D, no. 2, pp. 395-398, February 2010, doi: 10.1587/transinf.E93.D.395.
Abstract: To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used. For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination. We also investigate the robustness of matched and mismatched noise conditions between training and evaluation environments.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E93.D.395/_p
Copy
@ARTICLE{e93-d_2_395,
author={Makoto SAKAI, Norihide KITAOKA, Yuya HATTORI, Seiichi NAKAGAWA, Kazuya TAKEDA, },
journal={IEICE TRANSACTIONS on Information},
title={Evaluation of Combinational Use of Discriminant Analysis-Based Acoustic Feature Transformation and Discriminative Training},
year={2010},
volume={E93-D},
number={2},
pages={395-398},
abstract={To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used. For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination. We also investigate the robustness of matched and mismatched noise conditions between training and evaluation environments.},
keywords={},
doi={10.1587/transinf.E93.D.395},
ISSN={1745-1361},
month={February},}
Copy
TY - JOUR
TI - Evaluation of Combinational Use of Discriminant Analysis-Based Acoustic Feature Transformation and Discriminative Training
T2 - IEICE TRANSACTIONS on Information
SP - 395
EP - 398
AU - Makoto SAKAI
AU - Norihide KITAOKA
AU - Yuya HATTORI
AU - Seiichi NAKAGAWA
AU - Kazuya TAKEDA
PY - 2010
DO - 10.1587/transinf.E93.D.395
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E93-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2010
AB - To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used. For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination. We also investigate the robustness of matched and mismatched noise conditions between training and evaluation environments.
ER -