Initialization of fMLLR with Sufficient Statistics from Similar Speakers

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F11%3A43898196" target="_blank" >RIV/49777513:23520/11:43898196 - isvavai.cz</a>

Result on the web

<a href="http://dx.doi.org/10.1007/978-3-642-23538-2_24" target="_blank" >http://dx.doi.org/10.1007/978-3-642-23538-2_24</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-642-23538-2_24" target="_blank" >10.1007/978-3-642-23538-2_24</a>

Result language

angličtina

Original language name

Initialization of fMLLR with Sufficient Statistics from Similar Speakers

Original language description

One of the most utilized adaptation techniques is the feature Maximum Likelihood Linear Regression (fMLLR). In comparison with other adaptation methods the number of free parameters to be estimated significantly decreases. Thus, the method is well suitedfor situations with small amount of adaptation data. However, fMLLR still fails in situations with extremely small data sets. Such situations can be solved through proper initialization of fMLLR estimation adding some a-priori information. In this papera novel approach is proposed solving the problem of fMLLR initialization involving statistics from speakers acoustically close to the speaker to be adapted. Proposed initialization suitably substitutes missing adaptation data with similar data from a training database, fMLLR estimation becomes well-conditioned, and the accuracy of the recognition system increases even in situations with extremely small data sets.

Czech name

—

Czech description

—

Type

J_x - Unclassified - Peer-reviewed scientific article (Jimp, Jsc and Jost)

CEP classification

JD - Use of computers, robotics and its application

OECD FORD branch

—

Project

Result was created during the realization of more than one project. More information in the Projects tab.

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Publication year

2011

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Lecture Notes in Computer Science

ISSN

0302-9743

e-ISSN

—

Volume of the periodical

Neuveden

Issue of the periodical within the volume

6836

Country of publishing house

DE - GERMANY

Number of pages

8

Pages from-to

187-194

UT code for WoS article

—

EID of the result in the Scopus database

—