Skip to main navigation Skip to search Skip to main content

Evolutionary minimum verification error learning of the alternative hypothesis model for LLR-based speaker verification

  • Yi Hsiang Chao
  • , Wei Ho Tsai
  • , Shih Sian Cheng
  • , Hsin Min Wang
  • , Ruei Chuan Chang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

It is usually difficult to characterize the alternative hypothesis precisely in a log-likelihood ratio (LLR)-based speaker verification system. In a previous work, we proposed using a weighted arithmetic combination (WAC) or a weighted geometric combination (WGC) of the likelihoods of the background models instead of heuristic combinations, such as the arithmetic mean and the geometric mean, to better characterize the alternative hypothesis. In this paper, we further propose learning the parameters associated with WAC or WGC via an evolutionary minimum verification error (MVE) training method, such that both the false acceptance probability and the false rejection probability can be minimized. Our experiment results show that the proposed methods outperform conventional LLR-based approaches.

Original languageEnglish
Title of host publicationInternational Speech Communication Association - 8th Annual Conference of the International Speech Communication Association, Interspeech 2007
PublisherUnavailable
Pages1025-1028
Number of pages4
ISBN (Print)9781605603162
StatePublished - 2007
Event8th Annual Conference of the International Speech Communication Association, Interspeech 2007 - Antwerp, Belgium
Duration: 27 Aug 200731 Aug 2007

Publication series

NameInternational Speech Communication Association - 8th Annual Conference of the International Speech Communication Association, Interspeech 2007
Volume2
ISSN (Electronic)1990-9772

Conference

Conference8th Annual Conference of the International Speech Communication Association, Interspeech 2007
Country/TerritoryBelgium
CityAntwerp
Period27/08/0731/08/07

Keywords

  • Ggenetic algorithm
  • Log-likelihood ratio
  • Minimum verification error training
  • Speaker verification

Fingerprint

Dive into the research topics of 'Evolutionary minimum verification error learning of the alternative hypothesis model for LLR-based speaker verification'. Together they form a unique fingerprint.

Cite this