Search papers, labs, and topics across Lattice.
SphereVBx is a novel Bayesian clustering framework that utilizes Toroidal Probabilistic Spherical Discriminant Analysis (T-PSDA) to enhance speaker diarization through variational inference in a mixture of von Mises-Fisher distributions. By integrating this approach into the end-to-end neural diarization with vector clustering (EEND-VC) framework, the method not only simplifies the clustering process but also improves accuracy in various benchmarks. The parameter-free variant, SphereVBx-PF, further streamlines the model by eliminating the need for pretrained backend parameters while maintaining competitive performance.
SphereVBx achieves superior clustering accuracy in speaker diarization while significantly simplifying the process, making it a game-changer for EEND-VC applications.
We propose SphereVBx, a Bayesian clustering framework for hyperspherical embeddings based on Toroidal Probabilistic Spherical Discriminant Analysis (T-PSDA). The method follows the variational Bayesian formulation of VBx while replacing the Gaussian Probabilistic Linear Discriminant Analysis (PLDA) backend with T-PSDA, resulting in variational inference in a mixture of von Mises-Fisher distributions. We apply SphereVBx to speaker diarization and in particular to the end-to-end neural diarization with vector clustering (EEND-VC) framework. A parameter-free variant, denoted SphereVBx-PF, corresponds to a spherical similarity model closely related to cosine scoring and does not require pretrained backend parameters. Experiments on multiple diarization benchmarks show that SphereVBx improves clustering accuracy in cascaded diarization pipelines and achieves comparable or better performance in the EEND-VC framework while significantly simplifying its clustering stage.
