HMM-based speech synthesiser using the LF-model of the glottal source

J. Cabral; Steve Renals; J. Yamagishi; K. Richmond

doi:10.1109/ICASSP.2011.5947405

HMM-based speech synthesiser using the LF-model of the glottal source

J. Cabral, Steve Renals, J. Yamagishi, K. Richmond

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

Abstract

A major factor which causes a deterioration in speech quality in HMM-based speech synthesis is the use of a simple delta pulse signal to generate the excitation of voiced speech. This paper sets out a new approach to using an acoustic glottal source model in HMM-based synthesisers instead of the traditional pulse signal. The goal is to improve speech quality and to better model and transform voice characteristics. We have found the new method decreases buzziness and also improves prosodic modelling. A perceptual evaluation has supported this finding by showing a 55.6% preference for the new system, as against the baseline. This improvement, while not being as significant as we had initially expected, does encourage us to work on developing the proposed speech synthesiser further.

Original language	English
Title of host publication	Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on
Pages	4704-4707
Number of pages	4
ISBN (Electronic)	978-1-4577-0537-3
DOIs	https://doi.org/10.1109/ICASSP.2011.5947405
Publication status	Published - 1 May 2011
Event	ICASSP 2011 - 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) - Prague, Czech Republic, United Kingdom Duration: 22 May 2011 → 27 May 2011

Conference

Conference	ICASSP 2011 - 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Country/Territory	United Kingdom
Period	22/05/11 → 27/05/11

Access to Document

10.1109/ICASSP.2011.5947405

HMM-BASED SPEECH SYNTHESISER USING THE LF-MODELAccepted author manuscript, 231 KB

http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=5947405

ULTRAX: Ultrax: Real-time tongue tracking for speech therapy using ultrasound
Richmond, K. (Principal Investigator), Renals, S. (Co-investigator), Cleland, J. (Co-Investigator (External)) & Scobbie, J. M. (Co-Investigator (External))
EPSRC
1/02/11 → 31/07/14
Project: Research
EMIME: Effective multilingual interaction in mobile environments. RTD Linked to RE7006
King, S. (Principal Investigator) & Renals, S. (Co-investigator)
EU government bodies
1/03/08 → 28/02/11
Project: Research
Data-driven articulatory modelling: foundations for a new generation of speech synthesis
Renals, S. (Principal Investigator), King, S. (Co-investigator) & Richmond, K. (Co-investigator)
EPSRC
1/11/06 → 30/04/10
Project: Research

Cite this

@inproceedings{4b1df407fc2947f09cc5f53d0439958b,

title = "HMM-based speech synthesiser using the LF-model of the glottal source",

abstract = "A major factor which causes a deterioration in speech quality in HMM-based speech synthesis is the use of a simple delta pulse signal to generate the excitation of voiced speech. This paper sets out a new approach to using an acoustic glottal source model in HMM-based synthesisers instead of the traditional pulse signal. The goal is to improve speech quality and to better model and transform voice characteristics. We have found the new method decreases buzziness and also improves prosodic modelling. A perceptual evaluation has supported this finding by showing a 55.6% preference for the new system, as against the baseline. This improvement, while not being as significant as we had initially expected, does encourage us to work on developing the proposed speech synthesiser further.",

author = "J. Cabral and Steve Renals and J. Yamagishi and K. Richmond",

year = "2011",

month = may,

day = "1",

doi = "10.1109/ICASSP.2011.5947405",

language = "English",

isbn = "978-1-4577-0538-0",

pages = "4704--4707",

booktitle = "Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on",

note = "ICASSP 2011 - 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) ; Conference date: 22-05-2011 Through 27-05-2011",

}

Cabral, J, Renals, S, Yamagishi, J & Richmond, K 2011, HMM-based speech synthesiser using the LF-model of the glottal source. in Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on. pp. 4704-4707, ICASSP 2011 - 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), United Kingdom, 22/05/11. https://doi.org/10.1109/ICASSP.2011.5947405

TY - GEN

T1 - HMM-based speech synthesiser using the LF-model of the glottal source

AU - Cabral, J.

AU - Renals, Steve

AU - Yamagishi, J.

AU - Richmond, K.

PY - 2011/5/1

Y1 - 2011/5/1

N2 - A major factor which causes a deterioration in speech quality in HMM-based speech synthesis is the use of a simple delta pulse signal to generate the excitation of voiced speech. This paper sets out a new approach to using an acoustic glottal source model in HMM-based synthesisers instead of the traditional pulse signal. The goal is to improve speech quality and to better model and transform voice characteristics. We have found the new method decreases buzziness and also improves prosodic modelling. A perceptual evaluation has supported this finding by showing a 55.6% preference for the new system, as against the baseline. This improvement, while not being as significant as we had initially expected, does encourage us to work on developing the proposed speech synthesiser further.

AB - A major factor which causes a deterioration in speech quality in HMM-based speech synthesis is the use of a simple delta pulse signal to generate the excitation of voiced speech. This paper sets out a new approach to using an acoustic glottal source model in HMM-based synthesisers instead of the traditional pulse signal. The goal is to improve speech quality and to better model and transform voice characteristics. We have found the new method decreases buzziness and also improves prosodic modelling. A perceptual evaluation has supported this finding by showing a 55.6% preference for the new system, as against the baseline. This improvement, while not being as significant as we had initially expected, does encourage us to work on developing the proposed speech synthesiser further.

U2 - 10.1109/ICASSP.2011.5947405

DO - 10.1109/ICASSP.2011.5947405

M3 - Conference contribution

SN - 978-1-4577-0538-0

SP - 4704

EP - 4707

BT - Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on

T2 - ICASSP 2011 - 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Y2 - 22 May 2011 through 27 May 2011

ER -

HMM-based speech synthesiser using the LF-model of the glottal source

Abstract

Conference

Access to Document

Fingerprint

Projects

ULTRAX: Ultrax: Real-time tongue tracking for speech therapy using ultrasound

EMIME: Effective multilingual interaction in mobile environments. RTD Linked to RE7006

Data-driven articulatory modelling: foundations for a new generation of speech synthesis

Cite this