FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891926223> ?p ?o ?g. }

• W2891926223 endingPage "1286" @default.
• W2891926223 startingPage "1277" @default.
• W2891926223 abstract "In most of existing deep convolutional neural networks (CNNs) for classification, global average (first-order) pooling (GAP) has become a standard module to summarize activations of the last convolution layer as final representation for prediction. Recent researches show integration of higher-order pooling (HOP) methods clearly improves performance of deep CNNs. However, both GAP and existing HOP methods assume unimodal distributions, which cannot fully capture statistics of convolutional activations, limiting representation ability of deep CNNs, especially for samples with complex contents. To overcome the above limitation, this paper proposes a global Gated Mixture of Second-order Pooling (GM-SOP) method to further improve representation ability of deep CNNs. To this end, we introduce a sparsity-constrained gating mechanism and propose a novel parametric SOP as component of mixture model. Given a bank of SOP candidates, our method can adaptively choose Top-K (K > 1) candidates for each input sample through the sparsity-constrained gating module, and performs weighted sum of outputs of K selected candidates as representation of the sample. The proposed GM-SOP can flexibly accommodate a large number of personalized SOP candidates in an efficient way, leading to richer representations. The deep networks with our GM-SOP can be end-to-end trained, having potential to characterize complex, multi-modal distributions. The proposed method is evaluated on two large scale image benchmarks (i.e., downsampled ImageNet-1K and Places365), and experimental results show our GM-SOP is superior to its counterparts and achieves very competitive performance. The source code will be available at http://www.peihuali.org/GM-SOP." @default.
• W2891926223 created "2018-09-27" @default.
• W2891926223 creator A5010623154 @default.
• W2891926223 creator A5018318136 @default.
• W2891926223 creator A5029031269 @default.
• W2891926223 creator A5052272812 @default.
• W2891926223 creator A5070918242 @default.
• W2891926223 date "2018-01-01" @default.
• W2891926223 modified "2023-10-17" @default.
• W2891926223 title "Global Gated Mixture of Second-order Pooling for Improving Deep Convolutional Neural Networks" @default.
• W2891926223 cites W1483804921 @default.
• W2891926223 cites W1579271636 @default.
• W2891926223 cites W1598557217 @default.
• W2891926223 cites W1686810756 @default.
• W2891926223 cites W1909952827 @default.
• W2891926223 cites W2025653905 @default.
• W2891926223 cites W2046098009 @default.
• W2891926223 cites W2064675550 @default.
• W2891926223 cites W2069820356 @default.
• W2891926223 cites W2093608496 @default.
• W2891926223 cites W2095705004 @default.
• W2891926223 cites W2097117768 @default.
• W2891926223 cites W2108598243 @default.
• W2891926223 cites W2112759033 @default.
• W2891926223 cites W2119138423 @default.
• W2891926223 cites W2121407732 @default.
• W2891926223 cites W2129812935 @default.
• W2891926223 cites W2150884987 @default.
• W2891926223 cites W2179042386 @default.
• W2891926223 cites W2194775991 @default.
• W2891926223 cites W2204257188 @default.
• W2891926223 cites W2324076434 @default.
• W2891926223 cites W2401231614 @default.
• W2891926223 cites W2560052444 @default.
• W2891926223 cites W2601789736 @default.
• W2891926223 cites W2725513608 @default.
• W2891926223 cites W2732026016 @default.
• W2891926223 cites W2740620254 @default.
• W2891926223 cites W2741430497 @default.
• W2891926223 cites W2752965000 @default.
• W2891926223 cites W2768611535 @default.
• W2891926223 cites W2949117887 @default.
• W2891926223 cites W2952339051 @default.
• W2891926223 cites W2962761264 @default.
• W2891926223 cites W2963420686 @default.
• W2891926223 cites W2963446712 @default.
• W2891926223 cites W2964044082 @default.
• W2891926223 cites W3118608800 @default.
• W2891926223 hasPublicationYear "2018" @default.
• W2891926223 type Work @default.
• W2891926223 sameAs 2891926223 @default.
• W2891926223 citedByCount "2" @default.
• W2891926223 countsByYear W28919262232019 @default.
• W2891926223 crossrefType "proceedings-article" @default.
• W2891926223 hasAuthorship W2891926223A5010623154 @default.
• W2891926223 hasAuthorship W2891926223A5018318136 @default.
• W2891926223 hasAuthorship W2891926223A5029031269 @default.
• W2891926223 hasAuthorship W2891926223A5052272812 @default.
• W2891926223 hasAuthorship W2891926223A5070918242 @default.
• W2891926223 hasConcept C105795698 @default.
• W2891926223 hasConcept C108583219 @default.
• W2891926223 hasConcept C11413529 @default.
• W2891926223 hasConcept C117251300 @default.
• W2891926223 hasConcept C119857082 @default.
• W2891926223 hasConcept C153180895 @default.
• W2891926223 hasConcept C154945302 @default.
• W2891926223 hasConcept C17744445 @default.
• W2891926223 hasConcept C185592680 @default.
• W2891926223 hasConcept C198531522 @default.
• W2891926223 hasConcept C199539241 @default.
• W2891926223 hasConcept C2776359362 @default.
• W2891926223 hasConcept C33923547 @default.
• W2891926223 hasConcept C41008148 @default.
• W2891926223 hasConcept C43617362 @default.
• W2891926223 hasConcept C45347329 @default.
• W2891926223 hasConcept C50644808 @default.
• W2891926223 hasConcept C70437156 @default.
• W2891926223 hasConcept C81363708 @default.
• W2891926223 hasConcept C94625758 @default.
• W2891926223 hasConceptScore W2891926223C105795698 @default.
• W2891926223 hasConceptScore W2891926223C108583219 @default.
• W2891926223 hasConceptScore W2891926223C11413529 @default.
• W2891926223 hasConceptScore W2891926223C117251300 @default.
• W2891926223 hasConceptScore W2891926223C119857082 @default.
• W2891926223 hasConceptScore W2891926223C153180895 @default.
• W2891926223 hasConceptScore W2891926223C154945302 @default.
• W2891926223 hasConceptScore W2891926223C17744445 @default.
• W2891926223 hasConceptScore W2891926223C185592680 @default.
• W2891926223 hasConceptScore W2891926223C198531522 @default.
• W2891926223 hasConceptScore W2891926223C199539241 @default.
• W2891926223 hasConceptScore W2891926223C2776359362 @default.
• W2891926223 hasConceptScore W2891926223C33923547 @default.
• W2891926223 hasConceptScore W2891926223C41008148 @default.
• W2891926223 hasConceptScore W2891926223C43617362 @default.
• W2891926223 hasConceptScore W2891926223C45347329 @default.
• W2891926223 hasConceptScore W2891926223C50644808 @default.
• W2891926223 hasConceptScore W2891926223C70437156 @default.
• W2891926223 hasConceptScore W2891926223C81363708 @default.

• next