Matches in SemOpenAlex for { <https://semopenalex.org/work/W2566071968> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2566071968 endingPage "38" @default.
- W2566071968 startingPage "29" @default.
- W2566071968 abstract "Convolutional 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." @default.
- W2566071968 created "2017-01-06" @default.
- W2566071968 creator A5010831226 @default.
- W2566071968 creator A5011088033 @default.
- W2566071968 creator A5011886931 @default.
- W2566071968 creator A5035035680 @default.
- W2566071968 creator A5063276261 @default.
- W2566071968 creator A5072716284 @default.
- W2566071968 creator A5075755125 @default.
- W2566071968 creator A5076369490 @default.
- W2566071968 creator A5087109375 @default.
- W2566071968 date "2016-01-01" @default.
- W2566071968 modified "2023-10-18" @default.
- W2566071968 title "On the Representativeness of Convolutional Neural Networks Layers" @default.
- W2566071968 cites W1959000896 @default.
- W2566071968 cites W2062118960 @default.
- W2566071968 cites W2117539524 @default.
- W2566071968 cites W2155541015 @default.
- W2566071968 cites W2161381512 @default.
- W2566071968 cites W2483873654 @default.
- W2566071968 cites W2613678836 @default.
- W2566071968 cites W2919115771 @default.
- W2566071968 doi "https://doi.org/10.3233/978-1-61499-696-5-29" @default.
- W2566071968 hasPublicationYear "2016" @default.
- W2566071968 type Work @default.
- W2566071968 sameAs 2566071968 @default.
- W2566071968 citedByCount "0" @default.
- W2566071968 crossrefType "book-chapter" @default.
- W2566071968 hasAuthorship W2566071968A5010831226 @default.
- W2566071968 hasAuthorship W2566071968A5011088033 @default.
- W2566071968 hasAuthorship W2566071968A5011886931 @default.
- W2566071968 hasAuthorship W2566071968A5035035680 @default.
- W2566071968 hasAuthorship W2566071968A5063276261 @default.
- W2566071968 hasAuthorship W2566071968A5072716284 @default.
- W2566071968 hasAuthorship W2566071968A5075755125 @default.
- W2566071968 hasAuthorship W2566071968A5076369490 @default.
- W2566071968 hasAuthorship W2566071968A5087109375 @default.
- W2566071968 hasConcept C105795698 @default.
- W2566071968 hasConcept C111919701 @default.
- W2566071968 hasConcept C119857082 @default.
- W2566071968 hasConcept C153180895 @default.
- W2566071968 hasConcept C154945302 @default.
- W2566071968 hasConcept C157659113 @default.
- W2566071968 hasConcept C2778572836 @default.
- W2566071968 hasConcept C33923547 @default.
- W2566071968 hasConcept C37381756 @default.
- W2566071968 hasConcept C41008148 @default.
- W2566071968 hasConcept C41608201 @default.
- W2566071968 hasConcept C73555534 @default.
- W2566071968 hasConcept C81363708 @default.
- W2566071968 hasConcept C97931131 @default.
- W2566071968 hasConcept C98045186 @default.
- W2566071968 hasConceptScore W2566071968C105795698 @default.
- W2566071968 hasConceptScore W2566071968C111919701 @default.
- W2566071968 hasConceptScore W2566071968C119857082 @default.
- W2566071968 hasConceptScore W2566071968C153180895 @default.
- W2566071968 hasConceptScore W2566071968C154945302 @default.
- W2566071968 hasConceptScore W2566071968C157659113 @default.
- W2566071968 hasConceptScore W2566071968C2778572836 @default.
- W2566071968 hasConceptScore W2566071968C33923547 @default.
- W2566071968 hasConceptScore W2566071968C37381756 @default.
- W2566071968 hasConceptScore W2566071968C41008148 @default.
- W2566071968 hasConceptScore W2566071968C41608201 @default.
- W2566071968 hasConceptScore W2566071968C73555534 @default.
- W2566071968 hasConceptScore W2566071968C81363708 @default.
- W2566071968 hasConceptScore W2566071968C97931131 @default.
- W2566071968 hasConceptScore W2566071968C98045186 @default.
- W2566071968 hasLocation W25660719681 @default.
- W2566071968 hasOpenAccess W2566071968 @default.
- W2566071968 hasPrimaryLocation W25660719681 @default.
- W2566071968 hasRelatedWork W2260451253 @default.
- W2566071968 hasRelatedWork W2295702360 @default.
- W2566071968 hasRelatedWork W2734994004 @default.
- W2566071968 hasRelatedWork W2764024122 @default.
- W2566071968 hasRelatedWork W2795882061 @default.
- W2566071968 hasRelatedWork W2798280084 @default.
- W2566071968 hasRelatedWork W2803461564 @default.
- W2566071968 hasRelatedWork W2893797465 @default.
- W2566071968 hasRelatedWork W2901654183 @default.
- W2566071968 hasRelatedWork W2903693113 @default.
- W2566071968 hasRelatedWork W2908846145 @default.
- W2566071968 hasRelatedWork W2914967359 @default.
- W2566071968 hasRelatedWork W2919560393 @default.
- W2566071968 hasRelatedWork W2947392838 @default.
- W2566071968 hasRelatedWork W2952589086 @default.
- W2566071968 hasRelatedWork W3048675253 @default.
- W2566071968 hasRelatedWork W3133557063 @default.
- W2566071968 hasRelatedWork W3133584291 @default.
- W2566071968 hasRelatedWork W3179961982 @default.
- W2566071968 hasRelatedWork W3210922724 @default.
- W2566071968 isParatext "false" @default.
- W2566071968 isRetracted "false" @default.
- W2566071968 magId "2566071968" @default.
- W2566071968 workType "book-chapter" @default.