Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech

The automatic analysis of speech to detect affective states may improve the way users interact with electronic devices. However, the analysis only at the acoustic level could be not enough to determine the emotion of a user in a realistic scenario. In this paper we analyzed the spontaneous speech re...

Descripción completa

Detalles Bibliográficos
Autores: Iriondo Sanz, Ignasi, Planet García, Santiago
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2012
País:España
Institución:Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)
Repositorio:Recercat. Dipósit de la Recerca de Catalunya
OAI Identifier:oai:recercat.cat:20.500.14342/3399
Acceso en línea:http://hdl.handle.net/20.500.14342/3399
https://doi.org/10.9781/ijimai.2012.166
Access Level:acceso abierto
Palabra clave:Reconeixement automàtic de la parla
Processament de la parla
Lingüística computacional
62
id ES_4cd9a33fc9b7ae1c0fc19f3cbc5ec45d
oai_identifier_str oai:recercat.cat:20.500.14342/3399
network_acronym_str ES
network_name_str España
repository_id_str
spelling Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from SpeechIriondo Sanz, IgnasiPlanet García, SantiagoReconeixement automàtic de la parlaProcessament de la parlaLingüística computacional62The automatic analysis of speech to detect affective states may improve the way users interact with electronic devices. However, the analysis only at the acoustic level could be not enough to determine the emotion of a user in a realistic scenario. In this paper we analyzed the spontaneous speech recordings of the FAU Aibo Corpus at the acoustic and linguistic levels to extract two sets of features. The acoustic set was reduced by a greedy procedure selecting the most relevant features to optimize the learning stage. We compared two versions of this greedy selection algorithm by performing the search of the relevant features forwards and backwards. We experimented with three classification approaches: Nave-Bayes, a support vector machine and a logistic model tree, and two fusion schemes: decision-level fusion, merging the hard-decisions of the acoustic and linguistic classifiers by means of a decision tree; and feature-level fusion, concatenating both sets of features before the learning stage. Despite the low performance achieved by the linguistic data, a dramatic improvement was achieved after its combination with the acoustic information, improving the results achieved by this second modality on its own. The results achieved by the classifiers using the parameters merged at feature level outperformed the classification results of the decision-level fusion scheme, despite the simplicity of the scheme. Moreover, the extremely reduced set of acoustic features obtained by the greedy forward search selection algorithm improved the results provided by the full set.Universidad Internacional de La Rioja (UNIR)Universitat Ramon Llull. La Salle2012info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersion8 p.http://hdl.handle.net/20.500.14342/3399https://doi.org/10.9781/ijimai.2012.166reponame:Recercat. Dipósit de la Recerca de Catalunyainstname:Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)InglésInternational Journal of Interactive Multimedia and Artificial Intelligence, 2012, Vol. 1, No. 6 (Setembre)Attribution 4.0 International© L'autor/ahttp://creativecommons.org/licenses/by/4.0/info:eu-repo/semantics/openAccessoai:recercat.cat:20.500.14342/33992026-05-29T05:05:01Z
dc.title.none.fl_str_mv Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
title Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
spellingShingle Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
Iriondo Sanz, Ignasi
Reconeixement automàtic de la parla
Processament de la parla
Lingüística computacional
62
title_short Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
title_full Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
title_fullStr Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
title_full_unstemmed Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
title_sort Comparative Study on Feature Selection and Fusion Schemes for Emotion Recognition from Speech
dc.creator.none.fl_str_mv Iriondo Sanz, Ignasi
Planet García, Santiago
author Iriondo Sanz, Ignasi
author_facet Iriondo Sanz, Ignasi
Planet García, Santiago
author_role author
author2 Planet García, Santiago
author2_role author
dc.contributor.none.fl_str_mv Universitat Ramon Llull. La Salle
dc.subject.none.fl_str_mv Reconeixement automàtic de la parla
Processament de la parla
Lingüística computacional
62
topic Reconeixement automàtic de la parla
Processament de la parla
Lingüística computacional
62
description The automatic analysis of speech to detect affective states may improve the way users interact with electronic devices. However, the analysis only at the acoustic level could be not enough to determine the emotion of a user in a realistic scenario. In this paper we analyzed the spontaneous speech recordings of the FAU Aibo Corpus at the acoustic and linguistic levels to extract two sets of features. The acoustic set was reduced by a greedy procedure selecting the most relevant features to optimize the learning stage. We compared two versions of this greedy selection algorithm by performing the search of the relevant features forwards and backwards. We experimented with three classification approaches: Nave-Bayes, a support vector machine and a logistic model tree, and two fusion schemes: decision-level fusion, merging the hard-decisions of the acoustic and linguistic classifiers by means of a decision tree; and feature-level fusion, concatenating both sets of features before the learning stage. Despite the low performance achieved by the linguistic data, a dramatic improvement was achieved after its combination with the acoustic information, improving the results achieved by this second modality on its own. The results achieved by the classifiers using the parameters merged at feature level outperformed the classification results of the decision-level fusion scheme, despite the simplicity of the scheme. Moreover, the extremely reduced set of acoustic features obtained by the greedy forward search selection algorithm improved the results provided by the full set.
publishDate 2012
dc.date.none.fl_str_mv 2012
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/20.500.14342/3399
https://doi.org/10.9781/ijimai.2012.166
url http://hdl.handle.net/20.500.14342/3399
https://doi.org/10.9781/ijimai.2012.166
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv International Journal of Interactive Multimedia and Artificial Intelligence, 2012, Vol. 1, No. 6 (Setembre)
dc.rights.none.fl_str_mv Attribution 4.0 International
© L'autor/a
http://creativecommons.org/licenses/by/4.0/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv Attribution 4.0 International
© L'autor/a
http://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv 8 p.
dc.publisher.none.fl_str_mv Universidad Internacional de La Rioja (UNIR)
publisher.none.fl_str_mv Universidad Internacional de La Rioja (UNIR)
dc.source.none.fl_str_mv reponame:Recercat. Dipósit de la Recerca de Catalunya
instname:Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)
instname_str Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)
reponame_str Recercat. Dipósit de la Recerca de Catalunya
collection Recercat. Dipósit de la Recerca de Catalunya
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869407653826920448
score 15,81155

Similares en LA Referencia