2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2011
DOI: 10.1109/icassp.2011.5947358
|View full text |Cite
|
Sign up to set email alerts
|

Simplification and optimization of i-vector extraction

Abstract: 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. T… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
105
0

Year Published

2012
2012
2019
2019

Publication Types

Select...
5
2
1

Relationship

0
8

Authors

Journals

citations
Cited by 102 publications
(106 citation statements)
references
References 3 publications
1
105
0
Order By: Relevance
“…The UBM and the i-vector extractor are estimated from appropriate training corpora. Methods to train the i-vector extractor and estimate the i-vectors can be found in (Dehak et al, 2011;Glembek et al, 2011).…”
Section: The I-vector Representationmentioning
confidence: 99%
“…The UBM and the i-vector extractor are estimated from appropriate training corpora. Methods to train the i-vector extractor and estimate the i-vectors can be found in (Dehak et al, 2011;Glembek et al, 2011).…”
Section: The I-vector Representationmentioning
confidence: 99%
“…A standard i-vector extractor was implemented for Kaldi as well (see footnote 1 in Page 2), based on the baseline system described in [23].…”
Section: I-vector System Configurationmentioning
confidence: 99%
“…Most GMM techniques use some variation of joint factor analysis (JFA) [25]. An offshoot of JFA is the i-vector technique which does away with the channel part of the model and falls back toward a PCA approach [26]. See Section 5.1 for more on the i-vector approach.…”
Section: Channel Mismatchmentioning
confidence: 99%
“…Glembek, et al [26] provide simplifications to the formulation of the i-vectors to reduce the memory usage and to increase the speed of computing the vectors. Glembek, et al [26] also explore linear transformations using principal component analysis (PCA) and Heteroscedastic Linear Discriminant Analysis 4 (HLDA) [64] to achieve orthogonality of the components of the Gaussian mixture.…”
Section: The I-vector Model (Total Variability Space)mentioning
confidence: 99%
See 1 more Smart Citation