Matches in SemOpenAlex for { <https://semopenalex.org/work/W2922509574> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2922509574 abstract "In standard Convolutional Neural Networks (CNNs), the receptive fields of artificial neurons in each layer are designed to share the same size. It is well-known in the neuroscience community that the receptive field size of visual cortical neurons are modulated by the stimulus, which has been rarely considered in constructing CNNs. We propose a dynamic selection mechanism in CNNs that allows each neuron to adaptively adjust its receptive field size based on multiple scales of input information. A building block called Selective Kernel (SK) unit is designed, in which multiple branches with different kernel sizes are fused using softmax attention that is guided by the information in these branches. Different attentions on these branches yield different sizes of the effective receptive fields of neurons in the fusion layer. Multiple SK units are stacked to a deep network termed Selective Kernel Networks (SKNets). On the ImageNet and CIFAR benchmarks, we empirically show that SKNet outperforms the existing state-of-the-art architectures with lower model complexity. Detailed analyses show that the neurons in SKNet can capture target objects with different scales, which verifies the capability of neurons for adaptively adjusting their receptive field sizes according to the input. The code and models are available at https://github.com/implus/SKNet." @default.
- W2922509574 created "2019-03-22" @default.
- W2922509574 creator A5004579631 @default.
- W2922509574 creator A5062318228 @default.
- W2922509574 creator A5062687402 @default.
- W2922509574 creator A5069917952 @default.
- W2922509574 date "2019-06-01" @default.
- W2922509574 modified "2023-10-18" @default.
- W2922509574 title "Selective Kernel Networks" @default.
- W2922509574 cites W1530946731 @default.
- W2922509574 cites W1677182931 @default.
- W2922509574 cites W2031787885 @default.
- W2922509574 cites W2071550934 @default.
- W2922509574 cites W2097117768 @default.
- W2922509574 cites W2114567332 @default.
- W2922509574 cites W2115441154 @default.
- W2922509574 cites W2116360511 @default.
- W2922509574 cites W2117539524 @default.
- W2922509574 cites W2128272608 @default.
- W2922509574 cites W2135254996 @default.
- W2922509574 cites W2144764737 @default.
- W2922509574 cites W2147800946 @default.
- W2922509574 cites W2183341477 @default.
- W2922509574 cites W2194775991 @default.
- W2922509574 cites W2220384803 @default.
- W2922509574 cites W2302086703 @default.
- W2922509574 cites W2549139847 @default.
- W2922509574 cites W2601564443 @default.
- W2922509574 cites W2605344185 @default.
- W2922509574 cites W2778955544 @default.
- W2922509574 cites W2866634454 @default.
- W2922509574 cites W2883780447 @default.
- W2922509574 cites W2884030607 @default.
- W2922509574 cites W2884585870 @default.
- W2922509574 cites W2962850830 @default.
- W2922509574 cites W2962965405 @default.
- W2922509574 cites W2963030892 @default.
- W2922509574 cites W2963163009 @default.
- W2922509574 cites W2963446712 @default.
- W2922509574 cites W2963495494 @default.
- W2922509574 cites W2964350391 @default.
- W2922509574 doi "https://doi.org/10.1109/cvpr.2019.00060" @default.
- W2922509574 hasPublicationYear "2019" @default.
- W2922509574 type Work @default.
- W2922509574 sameAs 2922509574 @default.
- W2922509574 citedByCount "1148" @default.
- W2922509574 countsByYear W29225095742018 @default.
- W2922509574 countsByYear W29225095742019 @default.
- W2922509574 countsByYear W29225095742020 @default.
- W2922509574 countsByYear W29225095742021 @default.
- W2922509574 countsByYear W29225095742022 @default.
- W2922509574 countsByYear W29225095742023 @default.
- W2922509574 crossrefType "proceedings-article" @default.
- W2922509574 hasAuthorship W2922509574A5004579631 @default.
- W2922509574 hasAuthorship W2922509574A5062318228 @default.
- W2922509574 hasAuthorship W2922509574A5062687402 @default.
- W2922509574 hasAuthorship W2922509574A5069917952 @default.
- W2922509574 hasBestOaLocation W29225095742 @default.
- W2922509574 hasConcept C114614502 @default.
- W2922509574 hasConcept C153180895 @default.
- W2922509574 hasConcept C154945302 @default.
- W2922509574 hasConcept C188441871 @default.
- W2922509574 hasConcept C19071747 @default.
- W2922509574 hasConcept C33923547 @default.
- W2922509574 hasConcept C41008148 @default.
- W2922509574 hasConcept C50644808 @default.
- W2922509574 hasConcept C74193536 @default.
- W2922509574 hasConcept C81363708 @default.
- W2922509574 hasConceptScore W2922509574C114614502 @default.
- W2922509574 hasConceptScore W2922509574C153180895 @default.
- W2922509574 hasConceptScore W2922509574C154945302 @default.
- W2922509574 hasConceptScore W2922509574C188441871 @default.
- W2922509574 hasConceptScore W2922509574C19071747 @default.
- W2922509574 hasConceptScore W2922509574C33923547 @default.
- W2922509574 hasConceptScore W2922509574C41008148 @default.
- W2922509574 hasConceptScore W2922509574C50644808 @default.
- W2922509574 hasConceptScore W2922509574C74193536 @default.
- W2922509574 hasConceptScore W2922509574C81363708 @default.
- W2922509574 hasLocation W29225095741 @default.
- W2922509574 hasLocation W29225095742 @default.
- W2922509574 hasOpenAccess W2922509574 @default.
- W2922509574 hasPrimaryLocation W29225095741 @default.
- W2922509574 hasRelatedWork W2899027234 @default.
- W2922509574 hasRelatedWork W2962876041 @default.
- W2922509574 hasRelatedWork W2980176872 @default.
- W2922509574 hasRelatedWork W3090555870 @default.
- W2922509574 hasRelatedWork W3107204728 @default.
- W2922509574 hasRelatedWork W3108503355 @default.
- W2922509574 hasRelatedWork W3170224572 @default.
- W2922509574 hasRelatedWork W4220732972 @default.
- W2922509574 hasRelatedWork W4226420367 @default.
- W2922509574 hasRelatedWork W4287591324 @default.
- W2922509574 isParatext "false" @default.
- W2922509574 isRetracted "false" @default.
- W2922509574 magId "2922509574" @default.
- W2922509574 workType "article" @default.