Publication Details

A Noise Robust I-Vector Extractor Using Vector Taylor Series For Speaker Recognition

LEI, Y.; BURGET, L.; SCHEFFER, N. A Noise Robust I-Vector Extractor Using Vector Taylor Series For Speaker Recognition. Proceedings of ICASSP 2013. Vancouver: IEEE Signal Processing Society, 2013. p. 6788-6791. ISBN: 978-1-4799-0355-9.
Czech title
Extraktor I-vektorů využívající vektorovou Taylorovu řadu pro rozpoznávání mluvčího odolné vůči šumu
Type
conference paper
Language
English
Authors
Lei Yun
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Scheffer Nicolas
URL
Keywords

speaker recognition, Vector Taylor Series, ivector, noisy speaker verification, noise compensation

Abstract

This article describes a successfull adapation of the VTS approach to speaker recognition by proposing a new i-vector extraction framework.

Annotation

We propose a novel approach for noise-robust speaker recognition, where the model of distortions caused by additive and convolutive noises is integrated into the i-vector extraction framework. The model is based on a vector taylor series (VTS) approximation widely successful in noise robust speech recognition. The model allows for extracting "cleaned-up" i-vectors which can be used in a standard i-vector back end. We evaluate the proposed framework on the PRISM corpus, a NIST-SRE like corpus, where noisy conditions were created by artificially adding babble noises to clean speech segments. Results show that using VTS i-vectors present significant improvements in all noisy conditions compared to a state-of-theart baseline speaker recognition. More importantly, the proposed framework is robust to noise, as improvements are maintained when the system is trained on clean data.

Published
2013
Pages
6788–6791
Proceedings
Proceedings of ICASSP 2013
ISBN
978-1-4799-0355-9
Publisher
IEEE Signal Processing Society
Place
Vancouver
BibTeX 
@inproceedings{BUT103500,
  author="Yun {Lei} and Lukáš {Burget} and Nicolas {Scheffer}",
  title="A Noise Robust I-Vector Extractor Using Vector Taylor Series For Speaker Recognition",
  booktitle="Proceedings of ICASSP 2013",
  year="2013",
  pages="6788--6791",
  publisher="IEEE Signal Processing Society",
  address="Vancouver",
  isbn="978-1-4799-0355-9",
  url="http://www.fit.vutbr.cz/research/groups/speech/publi/2013/lei_icassp2013_0006788.pdf"
}
Back to top