Unsupervised Word Discovery from Speech using Bayesian Hierarchical Models

O. Walter, R. Haeb-Umbach, in: 38th German Conference on Pattern Recognition (GCPR 2016), 2016.

Conference Paper | English
Author
Abstract
In this paper we demonstrate an algorithm to learn words from speech using non-parametric Bayesian hierarchical models in an unsupervised setting. We exploit the assumption of a hierarchical structure of speech, namely the formation of spoken words as a sequence of phonemes. We employ the Nested Hierarchical Pitman-Yor Language Model, which allows an a priori unknown and possibly unlimited number of words. We assume the n-gram probabilities of words, the m-gram probabilities of phoneme sequences in words and the phoneme sequences of the words themselves as latent variables to be learned. We evaluate the algorithm on a cross language task using an existing speech recognizer trained on English speech to decode speech in the Xitsonga language supplied for the 2015 ZeroSpeech challenge. We apply the learning algorithm on the resulting phoneme graphs and achieve the highest token precision and F score compared to present systems.
Publishing Year
Proceedings Title
38th German Conference on Pattern Recognition (GCPR 2016)
LibreCat-ID

Cite this

Walter O, Haeb-Umbach R. Unsupervised Word Discovery from Speech using Bayesian Hierarchical Models. In: 38th German Conference on Pattern Recognition (GCPR 2016). ; 2016.
Walter, O., & Haeb-Umbach, R. (2016). Unsupervised Word Discovery from Speech using Bayesian Hierarchical Models. In 38th German Conference on Pattern Recognition (GCPR 2016).
@inproceedings{Walter_Haeb-Umbach_2016, title={Unsupervised Word Discovery from Speech using Bayesian Hierarchical Models}, booktitle={38th German Conference on Pattern Recognition (GCPR 2016)}, author={Walter, Oliver and Haeb-Umbach, Reinhold}, year={2016} }
Walter, Oliver, and Reinhold Haeb-Umbach. “Unsupervised Word Discovery from Speech Using Bayesian Hierarchical Models.” In 38th German Conference on Pattern Recognition (GCPR 2016), 2016.
O. Walter and R. Haeb-Umbach, “Unsupervised Word Discovery from Speech using Bayesian Hierarchical Models,” in 38th German Conference on Pattern Recognition (GCPR 2016), 2016.
Walter, Oliver, and Reinhold Haeb-Umbach. “Unsupervised Word Discovery from Speech Using Bayesian Hierarchical Models.” 38th German Conference on Pattern Recognition (GCPR 2016), 2016.

Link(s) to Main File(s)
Access Level
Restricted Closed Access
External material:
Supplementary Material
Description
Presentation

Export

Marked Publications

Open Data LibreCat

Search this title in

Google Scholar