Simplification and optimization of i-vector extraction

This paper introduces some simplifications to the i-vector speaker recognition systems. I-vector extraction as well as training of the i-vector extractor can be an expensive task both in terms of memory and speed. Under certain assumptions, the formulas for i-vector extraction-also used in i-vector...

Full description

Saved in:
Bibliographic Details
Main Authors: Glembek, Ondrej, Burget, Lukas, Matejka, Pavel, Karafiat, Martin, Kenny, Patrick
Format: Conference Proceeding
Language:English
Subjects:
Accuracy
Complexity theory
Covariance matrix
Feature extraction
HLDA
i-vectors
Joint Factor Analysis
NIST
PCA
speaker recognition
Speech
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper introduces some simplifications to the i-vector speaker recognition systems. I-vector extraction as well as training of the i-vector extractor can be an expensive task both in terms of memory and speed. Under certain assumptions, the formulas for i-vector extraction-also used in i-vector extractor training-can be simplified and lead to a faster and memory more efficient code. The first assumption is that the GMM component alignment is constant across utterances and is given by the UBM GMM weights. The second assumption is that the i-vector extractor matrix can be linearly transformed so that its per-Gaussian components are orthogonal. We use PCA and HLDA to estimate this transform.
ISSN:1520-6149
2379-190X
DOI:10.1109/ICASSP.2011.5947358