Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
652 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 15, NO. 2, FEBRUARY 2007 Robust Speaker Adaptation by Weighted
 

Summary: 652 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 15, NO. 2, FEBRUARY 2007
Robust Speaker Adaptation by Weighted
Model Averaging Based on the Minimum
Description Length Criterion
Xiaodong Cui, Member, IEEE, and Abeer Alwan, Senior Member, IEEE
Abstract--The maximum likelihood linear regression (MLLR)
technique is widely used in speaker adaptation due to its effec-
tiveness and computational advantages. When the adaptation data
are sparse, MLLR performance degrades because of unreliable pa-
rameter estimation. In this paper, a robust MLLR speaker adap-
tation approach via weighted model averaging is investigated. A
variety of transformation structures is first chosen and a general
form of maximum likelihood (ML) estimation of the structures is
given. The minimum description length (MDL) principle is applied
to account for the compromise between transformation granularity
and descriptive ability regarding the tying patterns of structured
transformations with a regression tree. Weighted model averaging
across the candidate structures is then performed based on the nor-
malized MDL scores. Experimental results show that this kind of
model averaging in combination with regression tree tying gives ro-

  

Source: Alwan, Abeer - Electrical Engineering Department, University of California at Los Angeles

 

Collections: Computer Technologies and Information Sciences