FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897596871> ?p ?o ?g. }

• W2897596871 abstract "We propose a random convolutional neural network to generate a feature space in which we study image classification and retrieval performance. Put briefly we apply random convolutional blocks followed by global average pooling to generate a new feature, and we repeat this k times to produce a k-dimensional feature space. This can be interpreted as partitioning the space of image patches with random hyperplanes which we formalize as a random depthwise convolutional neural network. In the network's final layer we perform image classification and retrieval with the linear support vector machine and k-nearest neighbor classifiers and study other empirical properties. We show that the ratio of image pixel distribution similarity across classes to within classes is higher in our network's final layer compared to the input space. When we apply the linear support vector machine for image classification we see that the accuracy is higher than if we were to train just the final layer of VGG16, ResNet18, and DenseNet40 with random weights. In the same setting we compare it to an unsupervised feature learning method and find our accuracy to be comparable on CIFAR10 but higher on CIFAR100 and STL10. We see that the accuracy is not far behind that of trained networks, particularly in the top-k setting. For example the top-2 accuracy of our network is near 90% on both CIFAR10 and a 10-class mini ImageNet, and 85% on STL10. We find that k-nearest neighbor gives a comparable precision on the Corel Princeton Image Similarity Benchmark than if we were to use the final layer of trained networks. As with other networks we find that our network fails to a black box attack even though we lack a gradient and use the sign activation. We highlight sensitivity of our network to background as a potential pitfall and an advantage. Overall our work pushes the boundary of what can be achieved with random weights." @default.
• W2897596871 created "2018-10-26" @default.
• W2897596871 creator A5083898401 @default.
• W2897596871 creator A5087990990 @default.
• W2897596871 date "2018-06-14" @default.
• W2897596871 modified "2023-09-23" @default.
• W2897596871 title "Image classification and retrieval with random depthwise signed convolutional neural networks" @default.
• W2897596871 cites W114517082 @default.
• W2897596871 cites W148066010 @default.
• W2897596871 cites W1529800609 @default.
• W2897596871 cites W1560724230 @default.
• W2897596871 cites W1686810756 @default.
• W2897596871 cites W180242331 @default.
• W2897596871 cites W2062118960 @default.
• W2897596871 cites W2095705004 @default.
• W2897596871 cites W2101234009 @default.
• W2897596871 cites W2105767494 @default.
• W2897596871 cites W2112796928 @default.
• W2897596871 cites W2117539524 @default.
• W2897596871 cites W2118585731 @default.
• W2897596871 cites W2118858186 @default.
• W2897596871 cites W2120480077 @default.
• W2897596871 cites W2123229215 @default.
• W2897596871 cites W2144161366 @default.
• W2897596871 cites W2163605009 @default.
• W2897596871 cites W2194775991 @default.
• W2897596871 cites W2286818309 @default.
• W2897596871 cites W2546302380 @default.
• W2897596871 cites W2556550110 @default.
• W2897596871 cites W2603766943 @default.
• W2897596871 cites W2618353736 @default.
• W2897596871 cites W2618907597 @default.
• W2897596871 cites W2746314669 @default.
• W2897596871 cites W2798991696 @default.
• W2897596871 cites W2962899986 @default.
• W2897596871 cites W2963446712 @default.
• W2897596871 cites W3098357269 @default.
• W2897596871 cites W3101584714 @default.
• W2897596871 cites W3118608800 @default.
• W2897596871 cites W78356000 @default.
• W2897596871 doi "https://doi.org/10.48550/arxiv.1806.05789" @default.
• W2897596871 hasPublicationYear "2018" @default.
• W2897596871 type Work @default.
• W2897596871 sameAs 2897596871 @default.
• W2897596871 citedByCount "0" @default.
• W2897596871 crossrefType "posted-content" @default.
• W2897596871 hasAuthorship W2897596871A5083898401 @default.
• W2897596871 hasAuthorship W2897596871A5087990990 @default.
• W2897596871 hasBestOaLocation W28975968711 @default.
• W2897596871 hasConcept C103278499 @default.
• W2897596871 hasConcept C113238511 @default.
• W2897596871 hasConcept C115961682 @default.
• W2897596871 hasConcept C12267149 @default.
• W2897596871 hasConcept C13280743 @default.
• W2897596871 hasConcept C138885662 @default.
• W2897596871 hasConcept C153180895 @default.
• W2897596871 hasConcept C154945302 @default.
• W2897596871 hasConcept C1667742 @default.
• W2897596871 hasConcept C185798385 @default.
• W2897596871 hasConcept C205649164 @default.
• W2897596871 hasConcept C2524010 @default.
• W2897596871 hasConcept C2776401178 @default.
• W2897596871 hasConcept C33923547 @default.
• W2897596871 hasConcept C41008148 @default.
• W2897596871 hasConcept C41895202 @default.
• W2897596871 hasConcept C68693459 @default.
• W2897596871 hasConcept C70437156 @default.
• W2897596871 hasConcept C75294576 @default.
• W2897596871 hasConcept C81363708 @default.
• W2897596871 hasConcept C83665646 @default.
• W2897596871 hasConceptScore W2897596871C103278499 @default.
• W2897596871 hasConceptScore W2897596871C113238511 @default.
• W2897596871 hasConceptScore W2897596871C115961682 @default.
• W2897596871 hasConceptScore W2897596871C12267149 @default.
• W2897596871 hasConceptScore W2897596871C13280743 @default.
• W2897596871 hasConceptScore W2897596871C138885662 @default.
• W2897596871 hasConceptScore W2897596871C153180895 @default.
• W2897596871 hasConceptScore W2897596871C154945302 @default.
• W2897596871 hasConceptScore W2897596871C1667742 @default.
• W2897596871 hasConceptScore W2897596871C185798385 @default.
• W2897596871 hasConceptScore W2897596871C205649164 @default.
• W2897596871 hasConceptScore W2897596871C2524010 @default.
• W2897596871 hasConceptScore W2897596871C2776401178 @default.
• W2897596871 hasConceptScore W2897596871C33923547 @default.
• W2897596871 hasConceptScore W2897596871C41008148 @default.
• W2897596871 hasConceptScore W2897596871C41895202 @default.
• W2897596871 hasConceptScore W2897596871C68693459 @default.
• W2897596871 hasConceptScore W2897596871C70437156 @default.
• W2897596871 hasConceptScore W2897596871C75294576 @default.
• W2897596871 hasConceptScore W2897596871C81363708 @default.
• W2897596871 hasConceptScore W2897596871C83665646 @default.
• W2897596871 hasLocation W28975968711 @default.
• W2897596871 hasLocation W28975968712 @default.
• W2897596871 hasOpenAccess W2897596871 @default.
• W2897596871 hasPrimaryLocation W28975968711 @default.
• W2897596871 hasRelatedWork W1533007923 @default.
• W2897596871 hasRelatedWork W2009083069 @default.
• W2897596871 hasRelatedWork W2056016498 @default.
• W2897596871 hasRelatedWork W2291847203 @default.
• W2897596871 hasRelatedWork W2319888919 @default.

• next