Improving the Slovak LVCSR Performance by Cluster-Sensitive Acoustic Model Retraining

Improving the Slovak LVCSR Performance by Cluster-Sensitive Acoustic Model Retraining

In this paper, we present a cluster-dependent adaptation approach for HMM-based acoustic models. The proposed approach employs clustering techniques to group the original training utterances into clusters with predefined number. The clustered speech data are intended to adapt an initially pre-traine...

Full description

Bibliographic Details
Main Authors: Peter Viszlay, Marek Ecegi, Josef Juhar
Format: Article
Language:English
Published: VSB-Technical University of Ostrava 2015-01-01
Series:Advances in Electrical and Electronic Engineering
Subjects:
acoustic model
adaptation
cluster analysis
reestimation
weighted mean vector.
Online Access:http://advances.utc.sk/index.php/AEEE/article/view/1448

Internet

http://advances.utc.sk/index.php/AEEE/article/view/1448

Similar Items