Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890886717> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W2890886717 abstract "We present a novel deep learning-based dehazing method using adaptive patch splits. Our method applies quad-tree decomposition to an input image, yielding multiple patches with adaptive sizes. Then, each patch is fed into a Convolutional Neural Network (CNN) and classified into a single transmission value, in which a transmission map comprises transmission values from all patches. Homogeneous regions in the image are typically decomposed into large patches. Thus the method can save computational cost. Non-homogeneous regions are divided into small patches, which helps preserve local details in a transmission map. To train CNN, we synthesize numerous hazy images from haze-free images. Experimental results demonstrate our method surpasses state- of-the-art deep learning based algorithms quantitatively and qualitatively." @default.
- W2890886717 created "2018-09-27" @default.
- W2890886717 creator A5045083125 @default.
- W2890886717 creator A5051281972 @default.
- W2890886717 creator A5059457064 @default.
- W2890886717 date "2018-10-01" @default.
- W2890886717 modified "2023-10-16" @default.
- W2890886717 title "Adaptive Patch Based Convolutional Neural Network for Robust Dehazing" @default.
- W2890886717 cites W1552073401 @default.
- W2890886717 cites W1979666307 @default.
- W2890886717 cites W2065002911 @default.
- W2890886717 cites W2081418206 @default.
- W2890886717 cites W2097900287 @default.
- W2890886717 cites W2123094559 @default.
- W2890886717 cites W2128254161 @default.
- W2890886717 cites W2157095036 @default.
- W2890886717 cites W2256362396 @default.
- W2890886717 cites W2779176852 @default.
- W2890886717 doi "https://doi.org/10.1109/icip.2018.8451252" @default.
- W2890886717 hasPublicationYear "2018" @default.
- W2890886717 type Work @default.
- W2890886717 sameAs 2890886717 @default.
- W2890886717 citedByCount "15" @default.
- W2890886717 countsByYear W28908867172019 @default.
- W2890886717 countsByYear W28908867172020 @default.
- W2890886717 countsByYear W28908867172021 @default.
- W2890886717 countsByYear W28908867172022 @default.
- W2890886717 countsByYear W28908867172023 @default.
- W2890886717 crossrefType "proceedings-article" @default.
- W2890886717 hasAuthorship W2890886717A5045083125 @default.
- W2890886717 hasAuthorship W2890886717A5051281972 @default.
- W2890886717 hasAuthorship W2890886717A5059457064 @default.
- W2890886717 hasConcept C108583219 @default.
- W2890886717 hasConcept C113174947 @default.
- W2890886717 hasConcept C114614502 @default.
- W2890886717 hasConcept C115961682 @default.
- W2890886717 hasConcept C134306372 @default.
- W2890886717 hasConcept C153180895 @default.
- W2890886717 hasConcept C154945302 @default.
- W2890886717 hasConcept C31972630 @default.
- W2890886717 hasConcept C33923547 @default.
- W2890886717 hasConcept C41008148 @default.
- W2890886717 hasConcept C50644808 @default.
- W2890886717 hasConcept C66882249 @default.
- W2890886717 hasConcept C761482 @default.
- W2890886717 hasConcept C76155785 @default.
- W2890886717 hasConcept C81363708 @default.
- W2890886717 hasConceptScore W2890886717C108583219 @default.
- W2890886717 hasConceptScore W2890886717C113174947 @default.
- W2890886717 hasConceptScore W2890886717C114614502 @default.
- W2890886717 hasConceptScore W2890886717C115961682 @default.
- W2890886717 hasConceptScore W2890886717C134306372 @default.
- W2890886717 hasConceptScore W2890886717C153180895 @default.
- W2890886717 hasConceptScore W2890886717C154945302 @default.
- W2890886717 hasConceptScore W2890886717C31972630 @default.
- W2890886717 hasConceptScore W2890886717C33923547 @default.
- W2890886717 hasConceptScore W2890886717C41008148 @default.
- W2890886717 hasConceptScore W2890886717C50644808 @default.
- W2890886717 hasConceptScore W2890886717C66882249 @default.
- W2890886717 hasConceptScore W2890886717C761482 @default.
- W2890886717 hasConceptScore W2890886717C76155785 @default.
- W2890886717 hasConceptScore W2890886717C81363708 @default.
- W2890886717 hasLocation W28908867171 @default.
- W2890886717 hasOpenAccess W2890886717 @default.
- W2890886717 hasPrimaryLocation W28908867171 @default.
- W2890886717 hasRelatedWork W2611989081 @default.
- W2890886717 hasRelatedWork W3029198973 @default.
- W2890886717 hasRelatedWork W3133861977 @default.
- W2890886717 hasRelatedWork W3167935049 @default.
- W2890886717 hasRelatedWork W3193565141 @default.
- W2890886717 hasRelatedWork W4226493464 @default.
- W2890886717 hasRelatedWork W4293226380 @default.
- W2890886717 hasRelatedWork W4312417841 @default.
- W2890886717 hasRelatedWork W4313906399 @default.
- W2890886717 hasRelatedWork W4375867731 @default.
- W2890886717 isParatext "false" @default.
- W2890886717 isRetracted "false" @default.
- W2890886717 magId "2890886717" @default.
- W2890886717 workType "article" @default.