Alleviating the small sample-size problem in i-vector based speaker verification

Wei Rao, Man Wai Mak

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

7 Citations (Scopus)

Abstract

This paper investigates the small sample-size problem in i-vector based speaker verification systems. The idea of i-vectors is to represent the characteristics of speakers in the factors of a factor analyzer. Because the factor loading matrix defines the possible speakerand channel-variability of i-vectors, it is important to suppress the unwanted channel variability. Linear discriminant analysis (LDA), within-class covariance normalization (WCCN), and probabilistic LDA are commonly used for such purpose. These methods, however, require training data comprising many speakers each providing sufficient recording sessions for good performance. Performance will suffer when the number of speakers and/or number of sessions per speaker are too small. This paper compares four approaches to addressing this small sample-size problem: (1) preprocessing the i-vectors by PCA before applying LDA (PCA+LDA), (2) replacing the matrix inverse in LDA by pseudo-inverse, (3) applying multi-way LDA by exploiting the microphone and speaker labels of the training data, and (4) increasing the matrix rank in LDA by generating more i-vectors using utterance partitioning. Results based on NIST 2010 SRE suggests that utterance partitioning performs the best, followed by multi-way LDA and PCA+LDA.
Original languageEnglish
Title of host publication2012 8th International Symposium on Chinese Spoken Language Processing, ISCSLP 2012
Pages335-339
Number of pages5
DOIs
Publication statusPublished - 1 Dec 2012
Event2012 8th International Symposium on Chinese Spoken Language Processing, ISCSLP 2012 - Hong Kong, Hong Kong
Duration: 5 Dec 20128 Dec 2012

Conference

Conference2012 8th International Symposium on Chinese Spoken Language Processing, ISCSLP 2012
Country/TerritoryHong Kong
CityHong Kong
Period5/12/128/12/12

Keywords

  • i-vectors
  • LDA
  • multi-way LDA
  • Speaker verification
  • utterance partitioning

ASJC Scopus subject areas

  • Language and Linguistics
  • Linguistics and Language

Cite this