Exponential family Fisher vector for image classification

Autores
Sanchez, Jorge Adrian; Redolfi, Javier Andrés
Año de publicación
2015
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
One of the fundamental problems in image classification is to devise models that allow us to relate the images to higher-level semantic concepts in an efficient and reliable way. A widely used approach consists on extracting local descriptors from the images and to summarize them into an image-level representation. Within this framework, the Fisher vector (FV) is one of the most robust signatures to date. In the FV, local descriptors are modeled as samples drawn from a mixture of Gaussian pdfs. An image is represented by a gradient vector characterizing the distributions of samples w.r.t. the model. Equipped with robust features like SIFT, the FV has shown state-of-the-art performance on different recognition problems. However, it is not clear how it should be applied when the feature space is clearly non-Euclidean, leading to heuristics that ignore the underlying structure of the space. In this paper we generalize the Gaussian FV to a broader family of distributions known as the exponential family. The model, termed exponential family Fisher vectors (eFV), provides a unified framework from which rich and powerful representations can be derived. Experimental results show the generality and flexibility of our approach.
Fil: Sanchez, Jorge Adrian. Universidad Nacional de Córdoba. Facultad de Matemática, Astronomía y Física; Argentina
Fil: Redolfi, Javier Andrés. Universidad Tecnológica Nacional. Facultad Regional Córdoba. Centro de Investigación en Informática para la Ingeniería; Argentina
Materia
EXPONENTIAL FAMILY
FISHER KERNEL
FISHER VECTORS
IMAGE CLASSIFICATION
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/59825

id CONICETDig_7dad8939b5153ea9e57891c7ce6c73a2
oai_identifier_str oai:ri.conicet.gov.ar:11336/59825
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Exponential family Fisher vector for image classificationSanchez, Jorge AdrianRedolfi, Javier AndrésEXPONENTIAL FAMILYFISHER KERNELFISHER VECTORSIMAGE CLASSIFICATIONhttps://purl.org/becyt/ford/2.2https://purl.org/becyt/ford/2One of the fundamental problems in image classification is to devise models that allow us to relate the images to higher-level semantic concepts in an efficient and reliable way. A widely used approach consists on extracting local descriptors from the images and to summarize them into an image-level representation. Within this framework, the Fisher vector (FV) is one of the most robust signatures to date. In the FV, local descriptors are modeled as samples drawn from a mixture of Gaussian pdfs. An image is represented by a gradient vector characterizing the distributions of samples w.r.t. the model. Equipped with robust features like SIFT, the FV has shown state-of-the-art performance on different recognition problems. However, it is not clear how it should be applied when the feature space is clearly non-Euclidean, leading to heuristics that ignore the underlying structure of the space. In this paper we generalize the Gaussian FV to a broader family of distributions known as the exponential family. The model, termed exponential family Fisher vectors (eFV), provides a unified framework from which rich and powerful representations can be derived. Experimental results show the generality and flexibility of our approach.Fil: Sanchez, Jorge Adrian. Universidad Nacional de Córdoba. Facultad de Matemática, Astronomía y Física; ArgentinaFil: Redolfi, Javier Andrés. Universidad Tecnológica Nacional. Facultad Regional Córdoba. Centro de Investigación en Informática para la Ingeniería; ArgentinaElsevier Science2015-07info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/59825Sanchez, Jorge Adrian; Redolfi, Javier Andrés; Exponential family Fisher vector for image classification; Elsevier Science; Pattern Recognition Letters; 59; 7-2015; 26-320167-86551872-7344CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S0167865515000811info:eu-repo/semantics/altIdentifier/doi/10.1016/j.patrec.2015.03.010info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-sa/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-10-15T15:13:11Zoai:ri.conicet.gov.ar:11336/59825instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-10-15 15:13:12.016CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Exponential family Fisher vector for image classification
title Exponential family Fisher vector for image classification
spellingShingle Exponential family Fisher vector for image classification
Sanchez, Jorge Adrian
EXPONENTIAL FAMILY
FISHER KERNEL
FISHER VECTORS
IMAGE CLASSIFICATION
title_short Exponential family Fisher vector for image classification
title_full Exponential family Fisher vector for image classification
title_fullStr Exponential family Fisher vector for image classification
title_full_unstemmed Exponential family Fisher vector for image classification
title_sort Exponential family Fisher vector for image classification
dc.creator.none.fl_str_mv Sanchez, Jorge Adrian
Redolfi, Javier Andrés
author Sanchez, Jorge Adrian
author_facet Sanchez, Jorge Adrian
Redolfi, Javier Andrés
author_role author
author2 Redolfi, Javier Andrés
author2_role author
dc.subject.none.fl_str_mv EXPONENTIAL FAMILY
FISHER KERNEL
FISHER VECTORS
IMAGE CLASSIFICATION
topic EXPONENTIAL FAMILY
FISHER KERNEL
FISHER VECTORS
IMAGE CLASSIFICATION
purl_subject.fl_str_mv https://purl.org/becyt/ford/2.2
https://purl.org/becyt/ford/2
dc.description.none.fl_txt_mv One of the fundamental problems in image classification is to devise models that allow us to relate the images to higher-level semantic concepts in an efficient and reliable way. A widely used approach consists on extracting local descriptors from the images and to summarize them into an image-level representation. Within this framework, the Fisher vector (FV) is one of the most robust signatures to date. In the FV, local descriptors are modeled as samples drawn from a mixture of Gaussian pdfs. An image is represented by a gradient vector characterizing the distributions of samples w.r.t. the model. Equipped with robust features like SIFT, the FV has shown state-of-the-art performance on different recognition problems. However, it is not clear how it should be applied when the feature space is clearly non-Euclidean, leading to heuristics that ignore the underlying structure of the space. In this paper we generalize the Gaussian FV to a broader family of distributions known as the exponential family. The model, termed exponential family Fisher vectors (eFV), provides a unified framework from which rich and powerful representations can be derived. Experimental results show the generality and flexibility of our approach.
Fil: Sanchez, Jorge Adrian. Universidad Nacional de Córdoba. Facultad de Matemática, Astronomía y Física; Argentina
Fil: Redolfi, Javier Andrés. Universidad Tecnológica Nacional. Facultad Regional Córdoba. Centro de Investigación en Informática para la Ingeniería; Argentina
description One of the fundamental problems in image classification is to devise models that allow us to relate the images to higher-level semantic concepts in an efficient and reliable way. A widely used approach consists on extracting local descriptors from the images and to summarize them into an image-level representation. Within this framework, the Fisher vector (FV) is one of the most robust signatures to date. In the FV, local descriptors are modeled as samples drawn from a mixture of Gaussian pdfs. An image is represented by a gradient vector characterizing the distributions of samples w.r.t. the model. Equipped with robust features like SIFT, the FV has shown state-of-the-art performance on different recognition problems. However, it is not clear how it should be applied when the feature space is clearly non-Euclidean, leading to heuristics that ignore the underlying structure of the space. In this paper we generalize the Gaussian FV to a broader family of distributions known as the exponential family. The model, termed exponential family Fisher vectors (eFV), provides a unified framework from which rich and powerful representations can be derived. Experimental results show the generality and flexibility of our approach.
publishDate 2015
dc.date.none.fl_str_mv 2015-07
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/59825
Sanchez, Jorge Adrian; Redolfi, Javier Andrés; Exponential family Fisher vector for image classification; Elsevier Science; Pattern Recognition Letters; 59; 7-2015; 26-32
0167-8655
1872-7344
CONICET Digital
CONICET
url http://hdl.handle.net/11336/59825
identifier_str_mv Sanchez, Jorge Adrian; Redolfi, Javier Andrés; Exponential family Fisher vector for image classification; Elsevier Science; Pattern Recognition Letters; 59; 7-2015; 26-32
0167-8655
1872-7344
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S0167865515000811
info:eu-repo/semantics/altIdentifier/doi/10.1016/j.patrec.2015.03.010
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier Science
publisher.none.fl_str_mv Elsevier Science
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1846083280027254784
score 13.22299