Parametric speech synthesis using local and global sparse Gaussian processes

Tomoki Koriyama; Takashi Nose; Takao Kobayashi

doi:10.1109/MLSP.2014.6958921

Parametric speech synthesis using local and global sparse Gaussian processes

Tomoki Koriyama, Takashi Nose, Takao Kobayashi

ENG - Communications Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper describes an application of Gaussian process regression (GPR) to parametric speech synthesis. GPR enables us to predict synthetic speech parameters by utilizing exemplars of training speech data directly without converting the acoustic features of training data into too small number of model parameters thanks to nonparametric Bayesian regression. However, GPR inherently requires high computational cost and resources. In this paper, to alleviate this problem, we incorporate local and global sparse Gaussian process approximation into the statistical speech synthesis framework, and investigate trade-off between computational cost and speech synthesis performance through experiments. Moreover, we examine the way of choosing pseudo data set used for the sparse GP approximation.

Original language	English
Title of host publication	IEEE International Workshop on Machine Learning for Signal Processing, MLSP
Editors	Tulay Adali, Jan Larsen, Mamadou Mboup, Eric Moreau
Publisher	IEEE Computer Society
ISBN (Electronic)	9781479936946
DOIs	https://doi.org/10.1109/MLSP.2014.6958921
Publication status	Published - 2014 Nov 14
Event	2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014 - Reims, France Duration: 2014 Sept 21 → 2014 Sept 24

Publication series

Name	IEEE International Workshop on Machine Learning for Signal Processing, MLSP
ISSN (Print)	2161-0363
ISSN (Electronic)	2161-0371

Conference

Conference	2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014
Country/Territory	France
City	Reims
Period	14/9/21 → 14/9/24

Keywords

Gaussian process regression
parametric speech synthesis
partially independent conditional (PIC) approximation

ASJC Scopus subject areas

Human-Computer Interaction
Signal Processing

Access to Document

10.1109/MLSP.2014.6958921

Cite this

Koriyama, T., Nose, T., & Kobayashi, T. (2014). Parametric speech synthesis using local and global sparse Gaussian processes. In T. Adali, J. Larsen, M. Mboup, & E. Moreau (Eds.), IEEE International Workshop on Machine Learning for Signal Processing, MLSP [6958921] (IEEE International Workshop on Machine Learning for Signal Processing, MLSP). IEEE Computer Society. https://doi.org/10.1109/MLSP.2014.6958921

Parametric speech synthesis using local and global sparse Gaussian processes. / Koriyama, Tomoki; Nose, Takashi; Kobayashi, Takao.
IEEE International Workshop on Machine Learning for Signal Processing, MLSP. ed. / Tulay Adali; Jan Larsen; Mamadou Mboup; Eric Moreau. IEEE Computer Society, 2014. 6958921 (IEEE International Workshop on Machine Learning for Signal Processing, MLSP).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Koriyama, T, Nose, T & Kobayashi, T 2014, Parametric speech synthesis using local and global sparse Gaussian processes. in T Adali, J Larsen, M Mboup & E Moreau (eds), IEEE International Workshop on Machine Learning for Signal Processing, MLSP., 6958921, IEEE International Workshop on Machine Learning for Signal Processing, MLSP, IEEE Computer Society, 2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014, Reims, France, 14/9/21. https://doi.org/10.1109/MLSP.2014.6958921

Koriyama T, Nose T, Kobayashi T. Parametric speech synthesis using local and global sparse Gaussian processes. In Adali T, Larsen J, Mboup M, Moreau E, editors, IEEE International Workshop on Machine Learning for Signal Processing, MLSP. IEEE Computer Society. 2014. 6958921. (IEEE International Workshop on Machine Learning for Signal Processing, MLSP). doi: 10.1109/MLSP.2014.6958921

Koriyama, Tomoki ; Nose, Takashi ; Kobayashi, Takao. / Parametric speech synthesis using local and global sparse Gaussian processes. IEEE International Workshop on Machine Learning for Signal Processing, MLSP. editor / Tulay Adali ; Jan Larsen ; Mamadou Mboup ; Eric Moreau. IEEE Computer Society, 2014. (IEEE International Workshop on Machine Learning for Signal Processing, MLSP).

@inproceedings{abe90585525b411c844f4614c4a3fa88,

title = "Parametric speech synthesis using local and global sparse Gaussian processes",

abstract = "This paper describes an application of Gaussian process regression (GPR) to parametric speech synthesis. GPR enables us to predict synthetic speech parameters by utilizing exemplars of training speech data directly without converting the acoustic features of training data into too small number of model parameters thanks to nonparametric Bayesian regression. However, GPR inherently requires high computational cost and resources. In this paper, to alleviate this problem, we incorporate local and global sparse Gaussian process approximation into the statistical speech synthesis framework, and investigate trade-off between computational cost and speech synthesis performance through experiments. Moreover, we examine the way of choosing pseudo data set used for the sparse GP approximation.",

keywords = "Gaussian process regression, parametric speech synthesis, partially independent conditional (PIC) approximation",

author = "Tomoki Koriyama and Takashi Nose and Takao Kobayashi",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014 ; Conference date: 21-09-2014 Through 24-09-2014",

year = "2014",

month = nov,

day = "14",

doi = "10.1109/MLSP.2014.6958921",

language = "English",

series = "IEEE International Workshop on Machine Learning for Signal Processing, MLSP",

publisher = "IEEE Computer Society",

editor = "Tulay Adali and Jan Larsen and Mamadou Mboup and Eric Moreau",

booktitle = "IEEE International Workshop on Machine Learning for Signal Processing, MLSP",

}

TY - GEN

T1 - Parametric speech synthesis using local and global sparse Gaussian processes

AU - Koriyama, Tomoki

AU - Nose, Takashi

AU - Kobayashi, Takao

PY - 2014/11/14

Y1 - 2014/11/14

N2 - This paper describes an application of Gaussian process regression (GPR) to parametric speech synthesis. GPR enables us to predict synthetic speech parameters by utilizing exemplars of training speech data directly without converting the acoustic features of training data into too small number of model parameters thanks to nonparametric Bayesian regression. However, GPR inherently requires high computational cost and resources. In this paper, to alleviate this problem, we incorporate local and global sparse Gaussian process approximation into the statistical speech synthesis framework, and investigate trade-off between computational cost and speech synthesis performance through experiments. Moreover, we examine the way of choosing pseudo data set used for the sparse GP approximation.

AB - This paper describes an application of Gaussian process regression (GPR) to parametric speech synthesis. GPR enables us to predict synthetic speech parameters by utilizing exemplars of training speech data directly without converting the acoustic features of training data into too small number of model parameters thanks to nonparametric Bayesian regression. However, GPR inherently requires high computational cost and resources. In this paper, to alleviate this problem, we incorporate local and global sparse Gaussian process approximation into the statistical speech synthesis framework, and investigate trade-off between computational cost and speech synthesis performance through experiments. Moreover, we examine the way of choosing pseudo data set used for the sparse GP approximation.

KW - Gaussian process regression

KW - parametric speech synthesis

KW - partially independent conditional (PIC) approximation

UR - http://www.scopus.com/inward/record.url?scp=84912544573&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84912544573&partnerID=8YFLogxK

U2 - 10.1109/MLSP.2014.6958921

DO - 10.1109/MLSP.2014.6958921

M3 - Conference contribution

AN - SCOPUS:84912544573

T3 - IEEE International Workshop on Machine Learning for Signal Processing, MLSP

BT - IEEE International Workshop on Machine Learning for Signal Processing, MLSP

A2 - Adali, Tulay

A2 - Larsen, Jan

A2 - Mboup, Mamadou

A2 - Moreau, Eric

PB - IEEE Computer Society

T2 - 2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014

Y2 - 21 September 2014 through 24 September 2014

ER -

Parametric speech synthesis using local and global sparse Gaussian processes

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this