Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation

Ignasi Iriondo, Santiago Planet, Francesc Alías, Joan Claudi Socoró, Elisa Martínez

Research output: A Conference proceeding or a Chapter in BookConference contribution

2 Citations (Scopus)

Abstract

This paper presents the validation of the expressive content of an acted corpus produced to be used in speech synthesis. The use of acted speech can be rather lacking in authenticity and therefore its expressiveness validation is required. The goal is to obtain an automatic classifier able to prune the bad utterances -with wrong expressiveness-. Firstly, a subjective test has been conducted with almost ten percent of the corpus utterances. Secondly, objective techniques have been carried out by means of automatic identification of emotions using different algorithms applied to statistical features computed over the speech prosody. The relationship between both evaluations is achieved by an attribute selection process guided by a metric that measures the matching between the misclassified utterances by the users and the automatic process. The experiments show that this approach can be useful to provide a subset of utterances with poor or wrong expressive content.

Original languageEnglish
Title of host publicationComputational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings
Subtitle of host publicationLecture Notes in Computer Science
Pages646-653
Number of pages8
Volume4507
DOIs
Publication statusPublished - 2007
Externally publishedYes
Event9th International Work-Conference on Artificial Neural Networks, IWANN 2007 - San Sebastian, Spain
Duration: 20 Jun 200722 Jun 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4507 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Work-Conference on Artificial Neural Networks, IWANN 2007
CountrySpain
CitySan Sebastian
Period20/06/0722/06/07

Fingerprint

Subjective Evaluation
Expressiveness
Speech synthesis
Prosody
Speech Synthesis
Classifiers
Percent
Emotions
Classifier
Attribute
Metric
Subset
Experiments
Evaluation
Experiment
Corpus
Speech

Cite this

Iriondo, I., Planet, S., Alías, F., Socoró, J. C., & Martínez, E. (2007). Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation. In Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings: Lecture Notes in Computer Science (Vol. 4507, pp. 646-653). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4507 LNCS). https://doi.org/10.1007/978-3-540-73007-1_78
Iriondo, Ignasi ; Planet, Santiago ; Alías, Francesc ; Socoró, Joan Claudi ; Martínez, Elisa. / Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation. Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings: Lecture Notes in Computer Science. Vol. 4507 2007. pp. 646-653 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{106e290325fe45d1ac2b211ecdd34193,
title = "Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation",
abstract = "This paper presents the validation of the expressive content of an acted corpus produced to be used in speech synthesis. The use of acted speech can be rather lacking in authenticity and therefore its expressiveness validation is required. The goal is to obtain an automatic classifier able to prune the bad utterances -with wrong expressiveness-. Firstly, a subjective test has been conducted with almost ten percent of the corpus utterances. Secondly, objective techniques have been carried out by means of automatic identification of emotions using different algorithms applied to statistical features computed over the speech prosody. The relationship between both evaluations is achieved by an attribute selection process guided by a metric that measures the matching between the misclassified utterances by the users and the automatic process. The experiments show that this approach can be useful to provide a subset of utterances with poor or wrong expressive content.",
keywords = "Expressive speech corpus",
author = "Ignasi Iriondo and Santiago Planet and Francesc Al{\'i}as and Socor{\'o}, {Joan Claudi} and Elisa Mart{\'i}nez",
year = "2007",
doi = "10.1007/978-3-540-73007-1_78",
language = "English",
isbn = "9783540730064",
volume = "4507",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "646--653",
booktitle = "Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings",

}

Iriondo, I, Planet, S, Alías, F, Socoró, JC & Martínez, E 2007, Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation. in Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings: Lecture Notes in Computer Science. vol. 4507, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4507 LNCS, pp. 646-653, 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, San Sebastian, Spain, 20/06/07. https://doi.org/10.1007/978-3-540-73007-1_78

Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation. / Iriondo, Ignasi; Planet, Santiago; Alías, Francesc; Socoró, Joan Claudi; Martínez, Elisa.

Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings: Lecture Notes in Computer Science. Vol. 4507 2007. p. 646-653 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4507 LNCS).

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation

AU - Iriondo, Ignasi

AU - Planet, Santiago

AU - Alías, Francesc

AU - Socoró, Joan Claudi

AU - Martínez, Elisa

PY - 2007

Y1 - 2007

N2 - This paper presents the validation of the expressive content of an acted corpus produced to be used in speech synthesis. The use of acted speech can be rather lacking in authenticity and therefore its expressiveness validation is required. The goal is to obtain an automatic classifier able to prune the bad utterances -with wrong expressiveness-. Firstly, a subjective test has been conducted with almost ten percent of the corpus utterances. Secondly, objective techniques have been carried out by means of automatic identification of emotions using different algorithms applied to statistical features computed over the speech prosody. The relationship between both evaluations is achieved by an attribute selection process guided by a metric that measures the matching between the misclassified utterances by the users and the automatic process. The experiments show that this approach can be useful to provide a subset of utterances with poor or wrong expressive content.

AB - This paper presents the validation of the expressive content of an acted corpus produced to be used in speech synthesis. The use of acted speech can be rather lacking in authenticity and therefore its expressiveness validation is required. The goal is to obtain an automatic classifier able to prune the bad utterances -with wrong expressiveness-. Firstly, a subjective test has been conducted with almost ten percent of the corpus utterances. Secondly, objective techniques have been carried out by means of automatic identification of emotions using different algorithms applied to statistical features computed over the speech prosody. The relationship between both evaluations is achieved by an attribute selection process guided by a metric that measures the matching between the misclassified utterances by the users and the automatic process. The experiments show that this approach can be useful to provide a subset of utterances with poor or wrong expressive content.

KW - Expressive speech corpus

UR - http://www.scopus.com/inward/record.url?scp=38049169335&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-73007-1_78

DO - 10.1007/978-3-540-73007-1_78

M3 - Conference contribution

SN - 9783540730064

VL - 4507

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 646

EP - 653

BT - Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings

ER -

Iriondo I, Planet S, Alías F, Socoró JC, Martínez E. Validation of an expressive speech corpus by mapping automatic classification to subjective evaluation. In Computational and Ambient Intelligence - 9th International Work-Conference on Artificial Neural Networks, IWANN 2007, Proceedings: Lecture Notes in Computer Science. Vol. 4507. 2007. p. 646-653. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-73007-1_78