Show simple item record

dc.contributor.authorGarcía Gasulla, Darío
dc.contributor.authorMoreno, Jonatan
dc.contributor.authorRamos-Pollan, Raúl
dc.contributor.authorCasadiegos Barrios, Romel
dc.contributor.authorBéjar Alonso, Javier
dc.contributor.authorCortés García, Claudio Ulises
dc.contributor.authorAyguadé Parra, Eduard
dc.contributor.authorLabarta Mancho, Jesús José
dc.contributor.authorSuzumura, Toyotaro
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament de Ciències de la Computació
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
dc.contributor.otherBarcelona Supercomputing Center
dc.date.accessioned2016-11-11T10:34:34Z
dc.date.available2016-11-11T10:34:34Z
dc.date.issued2016
dc.identifier.citationGarcía, D., Moreno, J., Ramos-Pollan, R., Barrios, R., Béjar, J., Cortés, C., Ayguadé, E., Labarta, J., Suzumura, T. On the representativeness of convolutional neural networks layers. A: "Artificial Intelligence Research and Development: proceedings of the 19th International Conference of the Catalan Association for Artificial Intelligence: Barcelona, Catalonia, Spain, October 19–21, 2016". IOS PRESS EBOOKS, 2016, p. 29-38.
dc.identifier.isbn978-1-61499-695-8
dc.identifier.urihttp://hdl.handle.net/2117/96526
dc.description.abstractConvolutional Neural Networks (CNN) are the most popular of deep network models due to their applicability and success in image processing. Although plenty of effort has been made in designing and training better discriminative CNNs, little is yet known about the internal features these models learn. Questions like, what specific knowledge is coded within CNN layers, and how can it be used for other purposes besides discrimination, remain to be answered. To advance in the resolution of these questions, in this work we extract features from CNN layers, building vector representations from CNN activations. The resultant vector embedding is used to represent first images and then known image classes. On those representations we perform an unsupervised clustering process, with the goal of studying the hidden semantics captured in the embedding space. Several abstract entities untaught to the network emerge in this process, effectively defining a taxonomy of knowledge as perceived by the CNN. We evaluate and interpret these sets using WordNet, while studying the different behaviours exhibited by the layers of a CNN model according to their depth. Our results indicate that, while top (i.e., deeper) layers provide the most representative space, low layers also define descriptive dimensions.
dc.description.sponsorshipThis work was partially supported by the IBM/BSC Technology Center for Supercomputing (Joint Study Agreement, No. W156463), by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project and by the Generalitat de Catalunya (contracts 2014-SGR-1051).
dc.format.extent10 p.
dc.language.isoeng
dc.publisherIOS PRESS EBOOKS
dc.subjectÀrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
dc.subject.lcshNeural networks (Computer science)
dc.subject.lcshImage processing
dc.subject.otherConvolutional neural networks
dc.subject.otherVisual embeddings
dc.subject.otherUnsupervised learning
dc.titleOn the representativeness of convolutional neural networks layers
dc.typePart of book or chapter of book
dc.subject.lemacXarxes neuronals (Informàtica)
dc.subject.lemacImatges -- Processament
dc.contributor.groupUniversitat Politècnica de Catalunya. KEMLG - Grup d'Enginyeria del Coneixement i Aprenentatge Automàtic
dc.contributor.groupUniversitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
dc.identifier.doi10.3233/978-1-61499-696-5-29
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttp://ebooks.iospress.nl/volumearticle/45307
dc.rights.accessOpen Access
local.identifier.drac19257322
dc.description.versionPostprint (author's final draft)
dc.relation.projectidinfo:eu-repo/grantAgreement/MINECO/1PE/TIN2015-65316-P
local.citation.authorGarcía, D.; Moreno, J.; Ramos-Pollan, R.; Barrios, R.; Béjar, J.; Cortés, C.; Ayguadé, E.; Labarta, J.; Suzumura, T.
local.citation.publicationNameArtificial Intelligence Research and Development: proceedings of the 19th International Conference of the Catalan Association for Artificial Intelligence: Barcelona, Catalonia, Spain, October 19–21, 2016
local.citation.startingPage29
local.citation.endingPage38


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder