Abstract
Table B.1 shows ore detailed results for different values of the static weight α. It can be seen that especially in the case of one frame the choice of an inappropriate value of α leads to a dramatic increase in WER. The closer α gets to one,which would correspond to standard MLLR,the bigger is the increase in WER.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
(2002). MLLR Results. In: Goronzy, S. (eds) Robust Adaptation to Non-Native Accents in Automatic Speech Recognition. Lecture Notes in Computer Science(), vol 2560. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-36290-8_12
Download citation
DOI: https://doi.org/10.1007/3-540-36290-8_12
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-00325-0
Online ISBN: 978-3-540-36290-6
eBook Packages: Springer Book Archive