Variational Inference for Acoustic Unit Discovery

详细信息    查看全文

• 作者：Lucas Ondel ; ^{iondel@fit.vutbr.cz" class="auth_mail" title="E-mail the corresponding author} ; Luka&scaron ; Burget ; Jan Černocký
• 关键词：Bayesian non-parametric ; Variational Bayes ; acoustic unit discovery
• 刊名：Procedia Computer Science
• 出版年：2016
• 出版时间：2016
• 年：2016
• 卷：81
• 期：Complete
• 页码：80-86
• 全文大小：162 K

文摘

Recently, several nonparametric Bayesian models have been proposed to automatically discover acoustic units in unlabeled data. Most of them are trained using various versions of the Gibbs Sampling (GS) method. In this work, we consider Variational Bayes (VB) as alternative inference process. Even though VB yields an approximate solution of the posterior distribution it can be easily parallelized which makes it more suitable for large database. Results show that, notwithstanding VB inference is an order of magnitude faster, it outperforms GS in terms of accuracy.