Presentation 2012-03-12
Fully Bayesian speaker clustering based on hierarchical structured Dirichlet process mixture model
Naohiro TAWARA, Tetsuji OGAWA, Shinji WATANABE, Atsushi NAKAMURA, Tetsunori KOBAYASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We proposed a novel speaker clustering method by estimating the structure of a fully Bayesian utterance generative model with a hierarchical structure. We defined the hierarchical generative model as a mixture of GMMs, which represent each speaker's distribution. We approximately estimated this model by introducing a sampling method because strict estimation of this model was infeasible. From speaker clustering experiments, we showed that the proposed method was effective to the data in which the number of utterances varied from speaker to speaker, while the conventional method caused significant degradation in clustering accuracy for these data.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Speaker clustering / Fully Bayesian model / Sampling / Utterance-oriented Dirichlet process mixture model
Paper # IBISML2011-90
Date of Issue

Conference Information
Committee IBISML
Conference Date 2012/3/5(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Information-Based Induction Sciences and Machine Learning (IBISML)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Fully Bayesian speaker clustering based on hierarchical structured Dirichlet process mixture model
Sub Title (in English)
Keyword(1) Speaker clustering
Keyword(2) Fully Bayesian model
Keyword(3) Sampling
Keyword(4) Utterance-oriented Dirichlet process mixture model
1st Author's Name Naohiro TAWARA
1st Author's Affiliation Waseda University()
2nd Author's Name Tetsuji OGAWA
2nd Author's Affiliation Waseda University
3rd Author's Name Shinji WATANABE
3rd Author's Affiliation Mitsubishi Electric Research Laboratories:NTT Comunication Science Laboratories
4th Author's Name Atsushi NAKAMURA
4th Author's Affiliation NTT Comunication Science Laboratories
5th Author's Name Tetsunori KOBAYASHI
5th Author's Affiliation Waseda University
Date 2012-03-12
Paper # IBISML2011-90
Volume (vol) vol.111
Number (no) 480
Page pp.pp.-
#Pages 8
Date of Issue