Matches in SemOpenAlex for { <https://semopenalex.org/work/W1640292247> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W1640292247 abstract "Convolutional Neural Networks (CNNs) have set the state-of-the-art in many computer vision tasks in recent years. For this type of model, it is common to have millions of parameters to train, commonly requiring large datasets. We investigate a method to transfer learning across different texture classification problems, using CNNs, in order to take advantage of this type of architecture to problems with smaller datasets. We use a Convolutional Neural Network trained on a source dataset (with lots of data) to project the data of a target dataset (with limited data) onto another feature space, and then train a classifier on top of this new representation. Our experiments show that this technique can achieve good results in tasks with small datasets, by leveraging knowledge learned from tasks with larger datasets. Testing the method on the the Brodatz-32 dataset, we achieved an accuracy of 97.04% - superior to models trained with popular texture descriptors, such as Local Binary Patterns and Gabor Filters, and increasing the accuracy by 6 percentage points compared to a CNN trained directly on the Brodatz-32 dataset. We also present a visual analysis of the projected dataset, showing that the data is projected to a space where samples from the same class are clustered together - suggesting that the features learned by the CNN in the source task are relevant for the target task." @default.
- W1640292247 created "2016-06-24" @default.
- W1640292247 creator A5025564604 @default.
- W1640292247 creator A5026340083 @default.
- W1640292247 creator A5038884704 @default.
- W1640292247 creator A5041613681 @default.
- W1640292247 date "2015-07-01" @default.
- W1640292247 modified "2023-10-16" @default.
- W1640292247 title "Transfer learning between texture classification tasks using Convolutional Neural Networks" @default.
- W1640292247 cites W2020681746 @default.
- W1640292247 cites W2022274350 @default.
- W1640292247 cites W2044465660 @default.
- W1640292247 cites W2044951190 @default.
- W1640292247 cites W2056451267 @default.
- W1640292247 cites W2071562807 @default.
- W1640292247 cites W2101198412 @default.
- W1640292247 cites W2112720808 @default.
- W1640292247 cites W2117251230 @default.
- W1640292247 cites W2145560214 @default.
- W1640292247 cites W2152187293 @default.
- W1640292247 cites W2153635508 @default.
- W1640292247 cites W2159666059 @default.
- W1640292247 cites W2161381512 @default.
- W1640292247 cites W2163005195 @default.
- W1640292247 cites W2165698076 @default.
- W1640292247 cites W2167181625 @default.
- W1640292247 cites W4231109964 @default.
- W1640292247 cites W64027530 @default.
- W1640292247 doi "https://doi.org/10.1109/ijcnn.2015.7280558" @default.
- W1640292247 hasPublicationYear "2015" @default.
- W1640292247 type Work @default.
- W1640292247 sameAs 1640292247 @default.
- W1640292247 citedByCount "20" @default.
- W1640292247 countsByYear W16402922472016 @default.
- W1640292247 countsByYear W16402922472017 @default.
- W1640292247 countsByYear W16402922472019 @default.
- W1640292247 countsByYear W16402922472020 @default.
- W1640292247 countsByYear W16402922472021 @default.
- W1640292247 countsByYear W16402922472022 @default.
- W1640292247 countsByYear W16402922472023 @default.
- W1640292247 crossrefType "proceedings-article" @default.
- W1640292247 hasAuthorship W1640292247A5025564604 @default.
- W1640292247 hasAuthorship W1640292247A5026340083 @default.
- W1640292247 hasAuthorship W1640292247A5038884704 @default.
- W1640292247 hasAuthorship W1640292247A5041613681 @default.
- W1640292247 hasConcept C108583219 @default.
- W1640292247 hasConcept C115961682 @default.
- W1640292247 hasConcept C119857082 @default.
- W1640292247 hasConcept C12267149 @default.
- W1640292247 hasConcept C150899416 @default.
- W1640292247 hasConcept C153180895 @default.
- W1640292247 hasConcept C154945302 @default.
- W1640292247 hasConcept C162324750 @default.
- W1640292247 hasConcept C175154964 @default.
- W1640292247 hasConcept C187736073 @default.
- W1640292247 hasConcept C2780451532 @default.
- W1640292247 hasConcept C41008148 @default.
- W1640292247 hasConcept C52622490 @default.
- W1640292247 hasConcept C53533937 @default.
- W1640292247 hasConcept C66905080 @default.
- W1640292247 hasConcept C75294576 @default.
- W1640292247 hasConcept C81363708 @default.
- W1640292247 hasConcept C87335442 @default.
- W1640292247 hasConcept C95623464 @default.
- W1640292247 hasConceptScore W1640292247C108583219 @default.
- W1640292247 hasConceptScore W1640292247C115961682 @default.
- W1640292247 hasConceptScore W1640292247C119857082 @default.
- W1640292247 hasConceptScore W1640292247C12267149 @default.
- W1640292247 hasConceptScore W1640292247C150899416 @default.
- W1640292247 hasConceptScore W1640292247C153180895 @default.
- W1640292247 hasConceptScore W1640292247C154945302 @default.
- W1640292247 hasConceptScore W1640292247C162324750 @default.
- W1640292247 hasConceptScore W1640292247C175154964 @default.
- W1640292247 hasConceptScore W1640292247C187736073 @default.
- W1640292247 hasConceptScore W1640292247C2780451532 @default.
- W1640292247 hasConceptScore W1640292247C41008148 @default.
- W1640292247 hasConceptScore W1640292247C52622490 @default.
- W1640292247 hasConceptScore W1640292247C53533937 @default.
- W1640292247 hasConceptScore W1640292247C66905080 @default.
- W1640292247 hasConceptScore W1640292247C75294576 @default.
- W1640292247 hasConceptScore W1640292247C81363708 @default.
- W1640292247 hasConceptScore W1640292247C87335442 @default.
- W1640292247 hasConceptScore W1640292247C95623464 @default.
- W1640292247 hasLocation W16402922471 @default.
- W1640292247 hasOpenAccess W1640292247 @default.
- W1640292247 hasPrimaryLocation W16402922471 @default.
- W1640292247 hasRelatedWork W2910954186 @default.
- W1640292247 hasRelatedWork W2981628807 @default.
- W1640292247 hasRelatedWork W3012393889 @default.
- W1640292247 hasRelatedWork W3176438653 @default.
- W1640292247 hasRelatedWork W3189091156 @default.
- W1640292247 hasRelatedWork W3193641238 @default.
- W1640292247 hasRelatedWork W4205999209 @default.
- W1640292247 hasRelatedWork W4285815841 @default.
- W1640292247 hasRelatedWork W4379875147 @default.
- W1640292247 hasRelatedWork W4386087993 @default.
- W1640292247 isParatext "false" @default.
- W1640292247 isRetracted "false" @default.
- W1640292247 magId "1640292247" @default.
- W1640292247 workType "article" @default.