FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2969890541> ?p ?o ?g. }

• W2969890541 abstract "Successful visual recognition networks benefit from aggregating information spanning from a wide range of scales. Previous research has investigated information fusion of connected layers or multiple branches in a block, seeking to strengthen the power of multi-scale representations. Despite their great successes, existing practices often allocate the neurons for each scale manually, and keep the same ratio in all aggregation blocks of an entire network, rendering suboptimal performance. In this paper, we propose to learn the neuron allocation for aggregating multi-scale information in different building blocks of a deep network. The most informative output neurons in each block are preserved while others are discarded, and thus neurons for multiple scales are competitively and adaptively allocated. Our scale aggregation network (ScaleNet) is constructed by repeating a scale aggregation (SA) block that concatenates feature maps at a wide range of scales. Feature maps for each scale are generated by a stack of downsampling, convolution and upsampling operations. The data-driven neuron allocation and SA block achieve strong representational power at the cost of considerably low computational complexity. The proposed ScaleNet, by replacing all 3×3 convolutions in ResNet with our SA blocks, achieves better performance than ResNet and its outstanding variants like ResNeXt and SE-ResNet, in the same computational complexity. On ImageNet classification, ScaleNets absolutely reduce the top-1 error rate of ResNets by 1.12 (101 layers) and 1.82 (50 layers). On COCO object detection, ScaleNets absolutely improve the mAP with backbone of ResNets by 3.6 and 4.6 on Faster-RCNN, respectively. Code and models are released on https://github.com/Eli-YiLi/ScaleNet." @default.
• W2969890541 created "2019-08-29" @default.
• W2969890541 creator A5037101612 @default.
• W2969890541 creator A5054299179 @default.
• W2969890541 creator A5057842914 @default.
• W2969890541 creator A5064273534 @default.
• W2969890541 date "2019-06-01" @default.
• W2969890541 modified "2023-10-14" @default.
• W2969890541 title "Data-Driven Neuron Allocation for Scale Aggregation Networks" @default.
• W2969890541 cites W1903029394 @default.
• W2969890541 cites W2011873184 @default.
• W2969890541 cites W2097117768 @default.
• W2969890541 cites W2135254996 @default.
• W2969890541 cites W2194775991 @default.
• W2969890541 cites W2288122362 @default.
• W2969890541 cites W2549139847 @default.
• W2969890541 cites W2560023338 @default.
• W2969890541 cites W2752782242 @default.
• W2969890541 cites W2962992847 @default.
• W2969890541 cites W2963145730 @default.
• W2969890541 cites W2963323244 @default.
• W2969890541 cites W2963363373 @default.
• W2969890541 cites W2963382930 @default.
• W2969890541 cites W2963446712 @default.
• W2969890541 cites W2964217527 @default.
• W2969890541 cites W639708223 @default.
• W2969890541 doi "https://doi.org/10.1109/cvpr.2019.01179" @default.
• W2969890541 hasPublicationYear "2019" @default.
• W2969890541 type Work @default.
• W2969890541 sameAs 2969890541 @default.
• W2969890541 citedByCount "19" @default.
• W2969890541 countsByYear W29698905412018 @default.
• W2969890541 countsByYear W29698905412019 @default.
• W2969890541 countsByYear W29698905412020 @default.
• W2969890541 countsByYear W29698905412021 @default.
• W2969890541 countsByYear W29698905412022 @default.
• W2969890541 countsByYear W29698905412023 @default.
• W2969890541 crossrefType "proceedings-article" @default.
• W2969890541 hasAuthorship W2969890541A5037101612 @default.
• W2969890541 hasAuthorship W2969890541A5054299179 @default.
• W2969890541 hasAuthorship W2969890541A5057842914 @default.
• W2969890541 hasAuthorship W2969890541A5064273534 @default.
• W2969890541 hasBestOaLocation W29698905412 @default.
• W2969890541 hasConcept C108583219 @default.
• W2969890541 hasConcept C110384440 @default.
• W2969890541 hasConcept C115961682 @default.
• W2969890541 hasConcept C121332964 @default.
• W2969890541 hasConcept C138885662 @default.
• W2969890541 hasConcept C153180895 @default.
• W2969890541 hasConcept C154945302 @default.
• W2969890541 hasConcept C205711294 @default.
• W2969890541 hasConcept C2524010 @default.
• W2969890541 hasConcept C2776401178 @default.
• W2969890541 hasConcept C2777210771 @default.
• W2969890541 hasConcept C2778755073 @default.
• W2969890541 hasConcept C2944601119 @default.
• W2969890541 hasConcept C33923547 @default.
• W2969890541 hasConcept C41008148 @default.
• W2969890541 hasConcept C41895202 @default.
• W2969890541 hasConcept C45347329 @default.
• W2969890541 hasConcept C50644808 @default.
• W2969890541 hasConcept C62520636 @default.
• W2969890541 hasConcept C80444323 @default.
• W2969890541 hasConceptScore W2969890541C108583219 @default.
• W2969890541 hasConceptScore W2969890541C110384440 @default.
• W2969890541 hasConceptScore W2969890541C115961682 @default.
• W2969890541 hasConceptScore W2969890541C121332964 @default.
• W2969890541 hasConceptScore W2969890541C138885662 @default.
• W2969890541 hasConceptScore W2969890541C153180895 @default.
• W2969890541 hasConceptScore W2969890541C154945302 @default.
• W2969890541 hasConceptScore W2969890541C205711294 @default.
• W2969890541 hasConceptScore W2969890541C2524010 @default.
• W2969890541 hasConceptScore W2969890541C2776401178 @default.
• W2969890541 hasConceptScore W2969890541C2777210771 @default.
• W2969890541 hasConceptScore W2969890541C2778755073 @default.
• W2969890541 hasConceptScore W2969890541C2944601119 @default.
• W2969890541 hasConceptScore W2969890541C33923547 @default.
• W2969890541 hasConceptScore W2969890541C41008148 @default.
• W2969890541 hasConceptScore W2969890541C41895202 @default.
• W2969890541 hasConceptScore W2969890541C45347329 @default.
• W2969890541 hasConceptScore W2969890541C50644808 @default.
• W2969890541 hasConceptScore W2969890541C62520636 @default.
• W2969890541 hasConceptScore W2969890541C80444323 @default.
• W2969890541 hasLocation W29698905411 @default.
• W2969890541 hasLocation W29698905412 @default.
• W2969890541 hasOpenAccess W2969890541 @default.
• W2969890541 hasPrimaryLocation W29698905411 @default.
• W2969890541 hasRelatedWork W1983892167 @default.
• W2969890541 hasRelatedWork W1991429770 @default.
• W2969890541 hasRelatedWork W2062399876 @default.
• W2969890541 hasRelatedWork W2281134365 @default.
• W2969890541 hasRelatedWork W2607795551 @default.
• W2969890541 hasRelatedWork W3155117723 @default.
• W2969890541 hasRelatedWork W4306309518 @default.
• W2969890541 hasRelatedWork W4310746709 @default.
• W2969890541 hasRelatedWork W4385574037 @default.
• W2969890541 hasRelatedWork W4386075645 @default.
• W2969890541 isParatext "false" @default.
• W2969890541 isRetracted "false" @default.
• W2969890541 magId "2969890541" @default.

• next