From single to multiple enrollment i-vectors: Practical PLDA scoring variants for speaker verification

Abstract

The availability of multiple utterances (and hence, i-vectors) for speaker enrollment brings up several alternatives for their utilization with probabilistic linear discriminant analysis (PLDA). This paper provides an overview of their effective utilization, from a practical viewpoint. We derive expressions for the evaluation of the likelihood ratio for the multi-enrollment case, with details on the computation of the required matrix inversions and determinants. The performance of five different scoring methods, and the effect of i-vector length normalization is compared experimentally. We conclude that length normalization is a useful technique for all but one of the scoring methods considered, and averaging i-vectors is the most effective out of the methods compared. We also study the application of multicondition training on the PLDA model. Our experiments indicate that multicondition training is more effective in estimating PLDA hyperparameters than it is for likelihood computation. Finally, we look at the effect of the configuration of the enrollment data on PLDA scoring, studying the properties of conditional dependence and number-of-enrollment-utterances per target speaker. Our experiments indicate that these properties affect the performance of the PLDA model. These results further support the conclusion that i-vector averaging is a simple and effective way to process multiple enrollment utterances.

Speaker verification based on the fusion of speech acoustics and inverted articulatory signals

“…Specifically, each speaker with each lexicon content password is considered as a class and different phrases from the same speaker are labeled with separate classes in the PLDA model training (Larcher et al, 2014b). Since there are three target utterances for each enrollment, we used the multiple enrollment PLDA scoring approach (Rajan et al, 2014; Liu et al, 2014). Finally, we simply employed the equal weighted summation fusion approach at the score level to further enhance the performance.…”

Optimization of data-driven filterbank for automatic speaker verification

“…The extracted i-vectors contain channel information. In order to compensate the effect of channel, probabilistic linear discriminant analysis (PLDA) is used to compute the similarity between i-vectors of enrollment and test [70]. We use Gaussian PLDA (GPLDA) in our experiment which models the within-class covariance by a full-rank matrix.…”

Anti-Spoofing for Text-Independent Speaker Verification: An Initial Database, Comparison of Countermeasures, and Human Performance

“…The 50 enrolment utterances were merged into 10 sessions (each being the concatenation of 5 utterances); either 1 or 10 of these sessions were used in enrolment, for the two enrolment scenarios. For PLDA, when using 10 enrolment sessions, ivectors were extracted from each session then averaged as suggested in [66]; for JFA, all features from all sessions 12 Available at: http://www.irisa.fr/metiss/guig/spro/ were merged. We denote the ASV systems with 5 enrolment utterances (presented as 1 session) as GMM-UBM-5, JFA-5 or PLDA-5 and those with 50 enrolment utterances (presented as 10 sessions) as GMM-UBM-50, JFA-50 or PLDA-50.…”