Publication detail

Variational Inference for Acoustic Unit Discovery

ONDEL YANG, L. BURGET, L. ČERNOCKÝ, J.

Original Title

Variational Inference for Acoustic Unit Discovery

Type

conference paper

Language

English

Original Abstract

In this article we proposed to train a nonparametric Bayesian model for automatic units discovery within the Variational Bayes framework. Besides simplifying the training scheme, this approach proves to be fast and yields better solution which makes it more suitable for big databases. However, despite the improvement observed, the model still have difficulties with the diversity of speech and tends to learn a large part of unwanted variability. The HMM model for speech segment is convenient but unrealistic and most likely, stronger model will be needed if one wants to achieve accurate automatic units discovery. We plan to extent the present work by using the VB inference with more complex models, as in13, and to gain leverage of Bayesian language models14 to further improve the accuracy of the discovered units.

Keywords

Bayesian non-parametric, Variational Bayes, acoustic unit discovery

Authors

ONDEL YANG, L.; BURGET, L.; ČERNOCKÝ, J.

Released

9. 7. 2016

Publisher

Elsevier Science

Location

Yogyakarta

ISBN

1877-0509

Periodical

Procedia Computer Science

Year of study

2016

Number

81

State

Kingdom of the Netherlands

Pages from

80

Pages to

86

Pages count

7

URL

http://www.sciencedirect.com/science/article/pii/S1877050916300473

BibTex

@inproceedings{BUT131006,
  author="Lucas Antoine Francois {Ondel} and Lukáš {Burget} and Jan {Černocký}",
  title="Variational Inference for Acoustic Unit Discovery",
  booktitle="Procedia Computer Science",
  year="2016",
  journal="Procedia Computer Science",
  volume="2016",
  number="81",
  pages="80--86",
  publisher="Elsevier Science",
  address="Yogyakarta",
  doi="10.1016/j.procs.2016.04.033",
  issn="1877-0509",
  url="http://www.sciencedirect.com/science/article/pii/S1877050916300473"
}