How to deal with multiple-targets in speaker identification systems?

Yaniv Zigel, Moshe Wasserblat

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

25 Scopus citations

Abstract

In open-set speaker identification systems a known phenomenon is that the false alarm (accept) error rate increases dramatically when increasing the number of registered speakers (models). In this paper, we demonstrate this phenomenon and suggest a solution using a new model-dependent score-normalization technique, called Top-norm. The Top-norm method was specifically developed to improve results of open-set speaker identification systems. Also, we suggest a score-normalization parameter adaptation technique. Experiments performed using speaker recognition corpora arc described and demonstrate that the new method outperforms other normalization methods.

Original languageEnglish
Title of host publicationIEEE Odyssey 2006
Subtitle of host publicationWorkshop on Speaker and Language Recognition
DOIs
StatePublished - 1 Dec 2006
Externally publishedYes
EventIEEE Odyssey 2006: Workshop on Speaker and Language Recognition - San Juan, Puerto Rico
Duration: 28 Jun 200630 Jun 2006

Publication series

NameIEEE Odyssey 2006: Workshop on Speaker and Language Recognition

Conference

ConferenceIEEE Odyssey 2006: Workshop on Speaker and Language Recognition
Country/TerritoryPuerto Rico
CitySan Juan
Period28/06/0630/06/06

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Signal Processing
  • Software

Fingerprint

Dive into the research topics of 'How to deal with multiple-targets in speaker identification systems?'. Together they form a unique fingerprint.

Cite this