Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Improving the Effectiveness of Speaker Verification Domain Adaptation With Inadequate In-Domain Data

Published

Author(s)

Bengt Borgstrom, Elliot Singer, Douglas Reynolds, Omid Sadjadi, Gregory A. Sanders

Abstract

This paper addresses speaker verification domain adaptation with inadequate in-domain data. Specifically, we explore he cases where in-domain data sets do not include speaker labels, contain speakers with few samples, or contain speakers with low channel diversity. Existing domain adaptation methods are reviewed, and their shortcomings are discussed. We derive an unsupervised version of fully Bayesian adaptation which reduces the reliance on rich in-domain data. When applied to domain adaptation with inadequate in-domain data, the proposed approach yields competitive results when the samples per speaker are reduced, and outperforms existing supervised methods when the channel diversity is low, even without requiring speaker labels. These results are validated on the SRE16, which uses a highly inadequate in-domain data set.
Proceedings Title
Interspeech 2017
Conference Dates
August 20-24, 2017
Conference Location
Stockholm, SE

Keywords

speaker verification, unsupervised domain adaptation, Bayesian adaptation

Citation

Borgstrom, B. , Singer, E. , Reynolds, D. , Sadjadi, O. and Sanders, G. (2017), Improving the Effectiveness of Speaker Verification Domain Adaptation With Inadequate In-Domain Data, Interspeech 2017, Stockholm, SE (Accessed May 28, 2024)

Issues

If you have any questions about this publication or are having problems accessing it, please contact reflib@nist.gov.

Created August 23, 2017, Updated October 12, 2021