PLDA in the i-supervector space for text-independent speaker verification

Jiang, Ye; Lee, Kong; Wang, Longbiao

doi:10.1186/preaccept-1667880097114310

Cited by 3 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Integer Linear Programming I-Vector (ILP IV) clustering [15] extracts an i-vector for each cluster and computes the distances among all of them (PLDA [16], cosine [17] or Mahalanobis [18]). ILP clustering was inspired by the k-medoids algorithm which choose k observations as class centers.…”

Section: Ilp IVmentioning

confidence: 99%

S4D: Speaker Diarization Toolkit in Python

Broux

Desnous²,

Larcher³

et al. 2018

Interspeech 2018

View full text Add to dashboard Cite

In this paper, we present S4D, a new open-source Python toolkit dedicated to speaker diarization. S4D provides various state-ofthe-art components and the possibility to easily develop end-toend diarization prototype systems. S4D offers a large panel of clustering, segmentation, scoring and visualization algorithms. S4D has been thought to be easily understood, installed, modified and used in order to allow fast transfers of diarization technologies to industry and facilitate development of new approaches. Examples, benchmarks on standard tasks and tutorials are provided in this paper. S4D is an extension of the opensource toolkit for speaker recognition: SIDEKIT.

show abstract

Section: Ilp IVmentioning

confidence: 99%

S4D: Speaker Diarization Toolkit in Python

Broux

Desnous²,

Larcher³

et al. 2018

Interspeech 2018

View full text Add to dashboard Cite

show abstract

“…where N (x|µ, Σ) represents a Gaussian in x with mean µ and covariance Σ. Here it's worth to notice that the mathematical relationship between DoJoBa and joint Bayesian [15] is analogous (not exactly) to that between joint PLDA [11] and PLDA [16]. Compared to joint PLDA, DoJoBa allows the data to determine the appropriate dimensionality of the low-rank speaker and text subspaces for maximal discrimination, as opposed to requiring heuristic manual selections.…”

Section: Double Joint Bayesianmentioning

confidence: 99%

Double Joint Bayesian Modeling of DNN Local I-Vector for Text Dependent Speaker Verification with Random Digit Strings

et al. 2018

View full text Add to dashboard Cite

Double joint Bayesian is a recently introduced analysis method that models and explores multiple information explicitly from the samples to improve the verification performance. It was recently applied to voice pass phrase verification, result in better results on text dependent speaker verification task. However little is known about its effectiveness in other challenging situations such as speaker verification for short, text-constrained test utterances, e.g. random digit strings. Contrary to conventional joint Bayesian method that cannot make full use of multi-view information, double joint Bayesian can incorporate both intra-speaker/digit and inter-speaker/digit variation, and calculated the likelihood to describe whether the features having all labels consistent or not. We show that double joint Bayesian outperforms conventional method on modeling DNN local (digit-dependent) i-vectors for speaker verification with random prompted digit strings. Since the strength of both double joint Bayesian and conventional DNN local i-vector appear complementary, the combination significantly outperforms either of its components.

show abstract

“…Secondly, dimensionality reduction is not possible. As D is a supervector-size diagonal matrix, the latent variable has to be of supervector length as well [14].…”

Section: The I-vector Paradigmmentioning

confidence: 99%

Quasi-Factorial Prior for i-vector Extraction

Chen

Lee²,

Dai

et al. 2015

IEEE Signal Process. Lett.

View full text Add to dashboard Cite

We analyze the i-vector extraction from the perspective of the prior distribution exerted on the mean supervector of Gaussian mixture model (GMM). To this end, we start off with the analysis of the subspace prior which leads to the compressed representation in the standard i-vector extraction. We then propose the use of quasi-factorial prior and show how it impacts the total variability space and its application for i-vector extraction. The quasi-factorial prior could be used in a standalone manner, or in combination with a subspace prior. In the latter context, we found that the performance of the standard i-vector can be greatly improved with the use of quasi-factorial prior followed by a subspace prior. This assertion is confirmed through experiments conducted on the NIST 2010 Speaker Recognition Evaluation (SRE10) dataset.

show abstract

PLDA in the i-supervector space for text-independent speaker verification

Cited by 3 publications

References 9 publications

S4D: Speaker Diarization Toolkit in Python

S4D: Speaker Diarization Toolkit in Python

Double Joint Bayesian Modeling of DNN Local I-Vector for Text Dependent Speaker Verification with Random Digit Strings

Quasi-Factorial Prior for i-vector Extraction

Contact Info

Product

Resources

About