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

Abstract

In this paper, we advocate the use of the uncompressed form of i-vector and depend on subspace modeling using probabilistic linear discriminant analysis (PLDA) in handling the speaker and session (or channel) variability. An i-vector is a low-dimensional vector containing both speaker and channel information acquired from a speech segment. When PLDA is used on an i-vector, dimension reduction is performed twice: first in the i-vector extraction process and second in the PLDA model. Keeping the full dimensionality of the i-vector in the i-supervector space for PLDA modeling and scoring would avoid unnecessary loss of information. We refer to the uncompressed i-vector as the i-supervector. The drawback in using the i-supervector with PLDA is the inversion of large matrices in the estimation of the full posterior distribution, which we show can be solved rather efficiently by portioning large matrices into smaller blocks. We also introduce the Gaussianized rank-norm, as an alternative to whitening, for feature normalization prior to PLDA modeling. We found that the i-supervector performs better during normalization. A better performance is obtained by combining the i-supervector and i-vector at the score level. Furthermore, we also analyze the computational complexity of the i-supervector system, compared with that of the i-vector, at four different stages of loading matrix estimation, posterior extraction, PLDA modeling, and PLDA scoring.

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

“…Experiments were carried out on the core task (short2-short3) of NIST SRE08 [42]. We use two well-known metrics in evaluating the performance, namely, equal error rate (EER) and minimum detection cost (MinDCF).…”

“…However, whitening can never be possible for i-supervector due to data scarcity. To this end, we advocate the use of a Gaussianized version of rank norm [34,41]. The i-supervector is processed element-wise with warping functions mapping each dimension to a standard Gaussian distribution (instead of uniform distribution as in rank norm).…”

“…One option is to estimate the subspaces in a decoupled manner, which might lead to suboptimal solution [12,24]. In [34], we showed that the joint estimation of subspaces can be accomplished by partitioning large matrices into smaller blocks, thereby making the inversion and the joint estimation feasible. In this study, we present the same approach with more detail and further refinement.…”

Unifying Probabilistic Linear Discriminant Analysis Variants in Biometric Authentication

“…The most detailed derivations are given in [16]. Our version was based on [14] and accelerated in a similar style as in [16]. The algorithm 1 summarizes it and the details are presented in the appendix A.…”

“…A number of solutions for this problem have been introduced. In [14], the authors utilize a special matrix structure of PLDA model and manually derive equations for the required matrix inversions. In [15], the authors proposed a special change of variables that lead to a diagonalized versions of the required matrices.…”