Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136246742> ?p ?o ?g. }
- W3136246742 abstract "In recent years, neural architecture search (NAS) methods have been proposed for the automatic generation of task-oriented network architecture in image classification. However, the architectures obtained by existing NAS approaches are optimized only for classification performance and do not adapt to devices with limited computational resources. To address this challenge, we propose a neural network architecture search algorithm aiming to simultaneously improve network performance (e.g., classification accuracy) and reduce network complexity. The proposed framework automatically builds the network architecture at two stages: block-level search and network-level search. At the stage of block-level search, a relaxation method based on the gradient is proposed, using an enhanced gradient to design high-performance and low-complexity blocks. At the stage of network-level search, we apply an evolutionary multi-objective algorithm to complete the automatic design from blocks to the target network. The experiment results demonstrate that our method outperforms all evaluated hand-crafted networks in image classification, with an error rate of on CIFAR10 and an error rate of on CIFAR100, both at network parameter size less than one megabit. Moreover, compared with other neural architecture search methods, our method offers a tremendous reduction in designed network architecture parameters." @default.
- W3136246742 created "2021-03-29" @default.
- W3136246742 creator A5001762132 @default.
- W3136246742 creator A5025035701 @default.
- W3136246742 creator A5048492871 @default.
- W3136246742 creator A5052624521 @default.
- W3136246742 creator A5064572247 @default.
- W3136246742 date "2021-03-23" @default.
- W3136246742 modified "2023-09-25" @default.
- W3136246742 title "Enhanced Gradient for Differentiable Architecture Search" @default.
- W3136246742 cites W1607307504 @default.
- W3136246742 cites W1653273292 @default.
- W3136246742 cites W1659842140 @default.
- W3136246742 cites W2097117768 @default.
- W3136246742 cites W2111935653 @default.
- W3136246742 cites W2112299196 @default.
- W3136246742 cites W2117539524 @default.
- W3136246742 cites W2119814172 @default.
- W3136246742 cites W2124659975 @default.
- W3136246742 cites W2126105956 @default.
- W3136246742 cites W2132194070 @default.
- W3136246742 cites W2145339207 @default.
- W3136246742 cites W2194775991 @default.
- W3136246742 cites W2266822037 @default.
- W3136246742 cites W2553303224 @default.
- W3136246742 cites W2606006859 @default.
- W3136246742 cites W2618530766 @default.
- W3136246742 cites W2782417188 @default.
- W3136246742 cites W2796265726 @default.
- W3136246742 cites W2806920391 @default.
- W3136246742 cites W2808938483 @default.
- W3136246742 cites W2810075754 @default.
- W3136246742 cites W2899243211 @default.
- W3136246742 cites W2916118939 @default.
- W3136246742 cites W2925665976 @default.
- W3136246742 cites W2953308748 @default.
- W3136246742 cites W2954070046 @default.
- W3136246742 cites W2963125010 @default.
- W3136246742 cites W2963163009 @default.
- W3136246742 cites W2963446712 @default.
- W3136246742 cites W2963778169 @default.
- W3136246742 cites W2963821229 @default.
- W3136246742 cites W2963918968 @default.
- W3136246742 cites W2964081807 @default.
- W3136246742 cites W2964212578 @default.
- W3136246742 cites W2965658867 @default.
- W3136246742 cites W2977955221 @default.
- W3136246742 cites W2981748264 @default.
- W3136246742 cites W3001340915 @default.
- W3136246742 cites W3002750555 @default.
- W3136246742 cites W3012028708 @default.
- W3136246742 cites W3034552520 @default.
- W3136246742 cites W3080443832 @default.
- W3136246742 cites W3082366738 @default.
- W3136246742 cites W3092580847 @default.
- W3136246742 doi "https://doi.org/10.48550/arxiv.2103.12529" @default.
- W3136246742 hasPublicationYear "2021" @default.
- W3136246742 type Work @default.
- W3136246742 sameAs 3136246742 @default.
- W3136246742 citedByCount "0" @default.
- W3136246742 crossrefType "posted-content" @default.
- W3136246742 hasAuthorship W3136246742A5001762132 @default.
- W3136246742 hasAuthorship W3136246742A5025035701 @default.
- W3136246742 hasAuthorship W3136246742A5048492871 @default.
- W3136246742 hasAuthorship W3136246742A5052624521 @default.
- W3136246742 hasAuthorship W3136246742A5064572247 @default.
- W3136246742 hasBestOaLocation W31362467421 @default.
- W3136246742 hasConcept C11413529 @default.
- W3136246742 hasConcept C115961682 @default.
- W3136246742 hasConcept C119857082 @default.
- W3136246742 hasConcept C123657996 @default.
- W3136246742 hasConcept C124101348 @default.
- W3136246742 hasConcept C125583679 @default.
- W3136246742 hasConcept C142362112 @default.
- W3136246742 hasConcept C153180895 @default.
- W3136246742 hasConcept C153349607 @default.
- W3136246742 hasConcept C154945302 @default.
- W3136246742 hasConcept C193415008 @default.
- W3136246742 hasConcept C2524010 @default.
- W3136246742 hasConcept C2777210771 @default.
- W3136246742 hasConcept C31258907 @default.
- W3136246742 hasConcept C33923547 @default.
- W3136246742 hasConcept C40969351 @default.
- W3136246742 hasConcept C41008148 @default.
- W3136246742 hasConcept C50644808 @default.
- W3136246742 hasConcept C75294576 @default.
- W3136246742 hasConceptScore W3136246742C11413529 @default.
- W3136246742 hasConceptScore W3136246742C115961682 @default.
- W3136246742 hasConceptScore W3136246742C119857082 @default.
- W3136246742 hasConceptScore W3136246742C123657996 @default.
- W3136246742 hasConceptScore W3136246742C124101348 @default.
- W3136246742 hasConceptScore W3136246742C125583679 @default.
- W3136246742 hasConceptScore W3136246742C142362112 @default.
- W3136246742 hasConceptScore W3136246742C153180895 @default.
- W3136246742 hasConceptScore W3136246742C153349607 @default.
- W3136246742 hasConceptScore W3136246742C154945302 @default.
- W3136246742 hasConceptScore W3136246742C193415008 @default.
- W3136246742 hasConceptScore W3136246742C2524010 @default.
- W3136246742 hasConceptScore W3136246742C2777210771 @default.
- W3136246742 hasConceptScore W3136246742C31258907 @default.