Gaussian meta-embeddings for efficient scoring of a heavy-tailed PLDA model
This work addresses uncertainty handling in speaker recognition, offering a domain-specific improvement over existing methods.
The authors tackled the problem of uncertainty propagation in speaker recognition by proposing Gaussian meta-embeddings with variable precisions from a heavy-tailed PLDA model, achieving up to 20% higher accuracy on NIST SRE datasets compared to the standard GPLDA method.
Embeddings in machine learning are low-dimensional representations of complex input patterns, with the property that simple geometric operations like Euclidean distances and dot products can be used for classification and comparison tasks. The proposed meta-embeddings are special embeddings that live in more general inner product spaces. They are designed to propagate uncertainty to the final output in speaker recognition and similar applications. The familiar Gaussian PLDA model (GPLDA) can be re-formulated as an extractor for Gaussian meta-embeddings (GMEs), such that likelihood ratio scores are given by Hilbert space inner products between Gaussian likelihood functions. GMEs extracted by the GPLDA model have fixed precisions and do not propagate uncertainty. We show that a generalization to heavy-tailed PLDA gives GMEs with variable precisions, which do propagate uncertainty. Experiments on NIST SRE 2010 and 2016 show that the proposed method applied to i-vectors without length normalization is up to 20% more accurate than GPLDA applied to length-normalized ivectors.