DMCA
Channel-dependent GMM and multi-class logistic regression models for language recognition (2006)
| Venue: | in Proceedings of the IEEE Odyssey 2006 Speaker and Language Recognition Workshop |
| Citations: | 5 - 0 self |
BibTeX
@INPROCEEDINGS{Leeuwen06channel-dependentgmm,
author = {David A. Van Leeuwen and Niko Brümmer and Spescom Datavoice and Stellenbosch South Africa},
title = {Channel-dependent GMM and multi-class logistic regression models for language recognition},
booktitle = {in Proceedings of the IEEE Odyssey 2006 Speaker and Language Recognition Workshop},
year = {2006}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This paper describes two new approaches to spoken language recognition. These were both successfully applied in the NIST 2005 Language Recognition Evaluation. The first approach extends the Gaussian Mixture Model technique with channel dependency, which results in actual detection costs (CDET) of 0.095 in NIST LRE-2005, and which should be compared to a traditional 2-gender dependency of GMM language models achieving 0.120. The second approach is a Multi-class Logistic Regression system, which operates similarly to a Support Vector Machine (SVM), but can be trained for all languages simultaneously. This new approach resulted in a CDET of 0.198. The joint TNO-Spescom Datavoice (TNO-SDV) submission to NIST LRE-2005 contained two more systems and obtained a result of 0.0958. 1.
Keyphrases
language recognition multi-class logistic regression model channel-dependent gmm nist lre-2005 new approach first approach joint tno-spescom datavoice second approach traditional 2-gender dependency language recognition evaluation support vector machine gaussian mixture model technique gmm language model channel dependency multi-class logistic regression system actual detection cost
