Adding personality to neutral speech synthesis voices

Christopher G. Buchanan; Matthew P. Aylett; David A. Braude

doi:10.1007/978-3-319-99579-3_6

Adding personality to neutral speech synthesis voices

Christopher G. Buchanan, Matthew P. Aylett, David A. Braude

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

3 Citations (Scopus)

Abstract

A synthetic voice personifies the system using it. Previous work has shown that using sub-corpora with different voice qualities (e.g. tense and lax) can be used to modify the perceived personality of a voice as well as adding expressive and emotional functionality. In this work we explore the use of LPC source/filter decomposition together with modification of the residual to artificially add voice quality sub-corpora to a voice without recording bespoke data. We evaluate this artificially enhanced voice against a baseline unit selection voice with pre-recorded sub-corpora. Although artificial modification impacts naturalness, it has the advantage of adding emotional range to voices where none was recorded in the source data, deals with data sparsity issues caused by sub-corpora, and results in significant effects in terms of perceived emotion.

Original language	English
Title of host publication	Speech and Computer. SPECOM 2018
Publisher	Springer
Pages	49-57
Number of pages	9
ISBN (Electronic)	978331999579-3
ISBN (Print)	9783319995786
DOIs	https://doi.org/10.1007/978-3-319-99579-3_6
Publication status	Published - 25 Aug 2018
Event	20th International Conference on Speech and Computer 2018 - Leipzig, Germany Duration: 18 Sept 2018 → 22 Sept 2018

Publication series

Name	Lecture Notes in Computer Science
Volume	11096
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th International Conference on Speech and Computer 2018
Abbreviated title	SPECOM 2018
Country/Territory	Germany
City	Leipzig
Period	18/09/18 → 22/09/18

Access to Document

10.1007/978-3-319-99579-3_6

Cite this

@inproceedings{afb0365798604dcdb9d5e14ebc4f9163,

title = "Adding personality to neutral speech synthesis voices",

abstract = "A synthetic voice personifies the system using it. Previous work has shown that using sub-corpora with different voice qualities (e.g. tense and lax) can be used to modify the perceived personality of a voice as well as adding expressive and emotional functionality. In this work we explore the use of LPC source/filter decomposition together with modification of the residual to artificially add voice quality sub-corpora to a voice without recording bespoke data. We evaluate this artificially enhanced voice against a baseline unit selection voice with pre-recorded sub-corpora. Although artificial modification impacts naturalness, it has the advantage of adding emotional range to voices where none was recorded in the source data, deals with data sparsity issues caused by sub-corpora, and results in significant effects in terms of perceived emotion.",

author = "Buchanan, {Christopher G.} and Aylett, {Matthew P.} and Braude, {David A.}",

year = "2018",

month = aug,

day = "25",

doi = "10.1007/978-3-319-99579-3_6",

language = "English",

isbn = "9783319995786",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "49--57",

booktitle = "Speech and Computer. SPECOM 2018",

note = "20th International Conference on Speech and Computer 2018, SPECOM 2018 ; Conference date: 18-09-2018 Through 22-09-2018",

}

TY - GEN

T1 - Adding personality to neutral speech synthesis voices

AU - Buchanan, Christopher G.

AU - Aylett, Matthew P.

AU - Braude, David A.

PY - 2018/8/25

Y1 - 2018/8/25

N2 - A synthetic voice personifies the system using it. Previous work has shown that using sub-corpora with different voice qualities (e.g. tense and lax) can be used to modify the perceived personality of a voice as well as adding expressive and emotional functionality. In this work we explore the use of LPC source/filter decomposition together with modification of the residual to artificially add voice quality sub-corpora to a voice without recording bespoke data. We evaluate this artificially enhanced voice against a baseline unit selection voice with pre-recorded sub-corpora. Although artificial modification impacts naturalness, it has the advantage of adding emotional range to voices where none was recorded in the source data, deals with data sparsity issues caused by sub-corpora, and results in significant effects in terms of perceived emotion.

AB - A synthetic voice personifies the system using it. Previous work has shown that using sub-corpora with different voice qualities (e.g. tense and lax) can be used to modify the perceived personality of a voice as well as adding expressive and emotional functionality. In this work we explore the use of LPC source/filter decomposition together with modification of the residual to artificially add voice quality sub-corpora to a voice without recording bespoke data. We evaluate this artificially enhanced voice against a baseline unit selection voice with pre-recorded sub-corpora. Although artificial modification impacts naturalness, it has the advantage of adding emotional range to voices where none was recorded in the source data, deals with data sparsity issues caused by sub-corpora, and results in significant effects in terms of perceived emotion.

U2 - 10.1007/978-3-319-99579-3_6

DO - 10.1007/978-3-319-99579-3_6

M3 - Conference contribution

SN - 9783319995786

T3 - Lecture Notes in Computer Science

SP - 49

EP - 57

BT - Speech and Computer. SPECOM 2018

PB - Springer

T2 - 20th International Conference on Speech and Computer 2018

Y2 - 18 September 2018 through 22 September 2018

ER -

Adding personality to neutral speech synthesis voices

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this