FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3041180677> ?p ?o ?g. }

• W3041180677 abstract "Replacing normal convolutions with group convolutions can significantly increase the computational efficiency of modern deep convolutional networks, which has been widely adopted in compact network architecture designs. However, existing group convolutions undermine the original network structures by cutting off some connections permanently resulting in significant accuracy degradation. In this paper, we propose dynamic group convolution (DGC) that adaptively selects which part of input channels to be connected within each group for individual samples on the fly. Specifically, we equip each group with a small feature selector to automatically select the most important input channels conditioned on the input images. Multiple groups can adaptively capture abundant and complementary visual/semantic features for each input image. The DGC preserves the original network structure and has similar computational efficiency as the conventional group convolution simultaneously. Extensive experiments on multiple image classification benchmarks including CIFAR-10, CIFAR-100 and ImageNet demonstrate its superiority over the existing group convolution techniques and dynamic execution methods. The code is available at https://github.com/zhuogege1943/dgc." @default.
• W3041180677 created "2020-07-16" @default.
• W3041180677 creator A5010794708 @default.
• W3041180677 creator A5014611768 @default.
• W3041180677 creator A5063481044 @default.
• W3041180677 creator A5070218966 @default.
• W3041180677 creator A5071074935 @default.
• W3041180677 creator A5089840086 @default.
• W3041180677 date "2020-07-08" @default.
• W3041180677 modified "2023-09-25" @default.
• W3041180677 title "Dynamic Group Convolution for Accelerating Convolutional Neural Networks" @default.
• W3041180677 cites W1861492603 @default.
• W3041180677 cites W1902041153 @default.
• W3041180677 cites W1903029394 @default.
• W3041180677 cites W2163605009 @default.
• W3041180677 cites W2167215970 @default.
• W3041180677 cites W2194775991 @default.
• W3041180677 cites W2515385951 @default.
• W3041180677 cites W2531409750 @default.
• W3041180677 cites W2549139847 @default.
• W3041180677 cites W2604998962 @default.
• W3041180677 cites W2612445135 @default.
• W3041180677 cites W2752037867 @default.
• W3041180677 cites W2778955544 @default.
• W3041180677 cites W2883780447 @default.
• W3041180677 cites W2884751099 @default.
• W3041180677 cites W2892446824 @default.
• W3041180677 cites W2894740066 @default.
• W3041180677 cites W2896006880 @default.
• W3041180677 cites W2906200245 @default.
• W3041180677 cites W2912168260 @default.
• W3041180677 cites W2924515500 @default.
• W3041180677 cites W2928560789 @default.
• W3041180677 cites W2945289329 @default.
• W3041180677 cites W2945908221 @default.
• W3041180677 cites W2948634622 @default.
• W3041180677 cites W2949117887 @default.
• W3041180677 cites W2950014519 @default.
• W3041180677 cites W2950967261 @default.
• W3041180677 cites W2951977814 @default.
• W3041180677 cites W2957777252 @default.
• W3041180677 cites W2961317571 @default.
• W3041180677 cites W2962851801 @default.
• W3041180677 cites W2962935523 @default.
• W3041180677 cites W2962944050 @default.
• W3041180677 cites W2963094099 @default.
• W3041180677 cites W2963125010 @default.
• W3041180677 cites W2963163009 @default.
• W3041180677 cites W2963363373 @default.
• W3041180677 cites W2963393494 @default.
• W3041180677 cites W2963403868 @default.
• W3041180677 cites W2963420686 @default.
• W3041180677 cites W2963446712 @default.
• W3041180677 cites W2963545688 @default.
• W3041180677 cites W2963674932 @default.
• W3041180677 cites W2963993763 @default.
• W3041180677 cites W2964062240 @default.
• W3041180677 cites W2964081807 @default.
• W3041180677 cites W2964169985 @default.
• W3041180677 cites W2965862350 @default.
• W3041180677 cites W2966256598 @default.
• W3041180677 cites W2968025890 @default.
• W3041180677 cites W2970601456 @default.
• W3041180677 cites W2970604398 @default.
• W3041180677 cites W2970657225 @default.
• W3041180677 cites W2970667965 @default.
• W3041180677 cites W2970860468 @default.
• W3041180677 cites W2970971581 @default.
• W3041180677 cites W2981819252 @default.
• W3041180677 cites W2981884310 @default.
• W3041180677 cites W2982157312 @default.
• W3041180677 cites W3118608800 @default.
• W3041180677 doi "https://doi.org/10.48550/arxiv.2007.04242" @default.
• W3041180677 hasPublicationYear "2020" @default.
• W3041180677 type Work @default.
• W3041180677 sameAs 3041180677 @default.
• W3041180677 citedByCount "3" @default.
• W3041180677 countsByYear W30411806772020 @default.
• W3041180677 countsByYear W30411806772021 @default.
• W3041180677 crossrefType "posted-content" @default.
• W3041180677 hasAuthorship W3041180677A5010794708 @default.
• W3041180677 hasAuthorship W3041180677A5014611768 @default.
• W3041180677 hasAuthorship W3041180677A5063481044 @default.
• W3041180677 hasAuthorship W3041180677A5070218966 @default.
• W3041180677 hasAuthorship W3041180677A5071074935 @default.
• W3041180677 hasAuthorship W3041180677A5089840086 @default.
• W3041180677 hasBestOaLocation W30411806771 @default.
• W3041180677 hasConcept C11413529 @default.
• W3041180677 hasConcept C115961682 @default.
• W3041180677 hasConcept C138885662 @default.
• W3041180677 hasConcept C153180895 @default.
• W3041180677 hasConcept C154945302 @default.
• W3041180677 hasConcept C177264268 @default.
• W3041180677 hasConcept C178790620 @default.
• W3041180677 hasConcept C185592680 @default.
• W3041180677 hasConcept C199360897 @default.
• W3041180677 hasConcept C2776401178 @default.
• W3041180677 hasConcept C2776760102 @default.
• W3041180677 hasConcept C2781311116 @default.
• W3041180677 hasConcept C41008148 @default.

• next