Publication detail
Fast variational Bayes for heavy-tailed PLDA applied to i-vectors and x-vectors
SILNOVA, A. BRUMMER, J. GARCÍA-ROMERO, D. SNYDER, D. BURGET, L.
Original Title
Fast variational Bayes for heavy-tailed PLDA applied to i-vectors and x-vectors
Type
conference paper
Language
English
Original Abstract
The standard state-of-the-art backend for text-independentspeaker recognizers that use i-vectors or x-vectors, is GaussianPLDA (G-PLDA), assisted by a Gaussianization step involvinglength normalization. G-PLDA can be trained withboth generative or discriminative methods. It has long beenknown that heavy-tailed PLDA (HT-PLDA), applied withoutlength normalization, gives similar accuracy, but at considerableextra computational cost. We have recently introduced afast scoring algorithm for a discriminatively trained HT-PLDAbackend. This paper extends that work by introducing a fast,variational Bayes, generative training algorithm. We compareold and new backends, with and without length-normalization,with i-vectors and x-vectors, on SRE10, SRE16 and SITW.
Keywords
peaker recognition, variational Bayes, heavytailed PLDA
Authors
SILNOVA, A.; BRUMMER, J.; GARCÍA-ROMERO, D.; SNYDER, D.; BURGET, L.
Released
2. 9. 2018
Publisher
International Speech Communication Association
Location
Hyderabad
ISBN
1990-9772
Periodical
Proceedings of Interspeech
Year of study
2018
Number
9
State
French Republic
Pages from
72
Pages to
76
Pages count
5
URL
BibTex
@inproceedings{BUT155098,
author="SILNOVA, A. and BRUMMER, J. and GARCÍA-ROMERO, D. and SNYDER, D. and BURGET, L.",
title="Fast variational Bayes for heavy-tailed PLDA applied to i-vectors and x-vectors",
booktitle="Proceedings of Interspeech 2018",
year="2018",
journal="Proceedings of Interspeech",
volume="2018",
number="9",
pages="72--76",
publisher="International Speech Communication Association",
address="Hyderabad",
doi="10.21437/Interspeech.2018-2128",
issn="1990-9772",
url="https://www.isca-speech.org/archive/Interspeech_2018/abstracts/2128.html"
}Documents