MLNov 20, 2015
Bayesian SPLDA
arXiv:1511.07318v1
Originality Synthesis-oriented
AI Analysis
This work addresses the problem of adapting speaker recognition models with minimal data, but it is incremental as it builds on existing methods.
The authors derived equations for Variational Bayes estimation of SPLDA parameters to enable adaptation between databases with limited data or implement a fully Bayesian approach, similar to Bishop's VB PPCA.
In this document we are going to derive the equations needed to implement a Variational Bayes estimation of the parameters of the simplified probabilistic linear discriminant analysis (SPLDA) model. This can be used to adapt SPLDA from one database to another with few development data or to implement the fully Bayesian recipe. Our approach is similar to Bishop's VB PPCA.