FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3161926058> ?p ?o ?g. }

• W3161926058 endingPage "41563" @default.
• W3161926058 startingPage "41551" @default.
• W3161926058 abstract "Deep Neural Networks (DNNs) are computationally and memory intensive, which present a big challenge for hardware, especially for resource-constrained devices such as Internet-of-Things (IoT) nodes. This paper introduces a new method to improve DNNs performance by fusing approximate computing with data reuse techniques for image recognition applications. First, starting from the pre-trained network, then the DNNs weights are approximated based on the linear and quadratic approximation methods during the retraining phase to reduce the DNN model size and number of arithmetic operations. Then, the DNNs weights are replaced with the linear/quadratic coefficients to execute the inference so that different DNNs weights can be computed using the same coefficients. That leads to a repetition of the weights, which enables the reuse of the DNN sub-computations (computational reuse) and leverages the same data (data reuse) to reduce DNNs computations memory accesses, and improve energy efficiency, albeit at the cost of increased training time. Complete analysis for MNIST, Fashion MNIST, CIFAR 10, CIFAR 100, and tiny ImageNet datasets is presented for image recognition, where different DNN models are used, including LeNet, ResNet, AlexNet, and VGG16. Our results show that the linear approximation achieves <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$1211.3times $ </tex-math></inline-formula> , <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$21.8times $ </tex-math></inline-formula> , <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$700times $ </tex-math></inline-formula> , and <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$19.3times $ </tex-math></inline-formula> on LeNet-5 MNIST, LeNet Fashion MNIST, VGG16 and ResNet-20. respectively, with small accuracy loss. Compared to the state-of-the-art Row Stationary (RS) method, the proposed architecture saved 54% of the total number of adders and multipliers needed. Overall, the proposed approach is suitable for IoT edge devices as it reduces computing complexity, memory size, and memory access with a small impact on accuracy." @default.
• W3161926058 created "2021-05-24" @default.
• W3161926058 creator A5013814572 @default.
• W3161926058 creator A5020682177 @default.
• W3161926058 creator A5062311050 @default.
• W3161926058 creator A5078108496 @default.
• W3161926058 creator A5091151249 @default.
• W3161926058 date "2022-01-01" @default.
• W3161926058 modified "2023-09-25" @default.
• W3161926058 title "Deep Neural Networks-Based Weight Approximation and Computation Reuse for 2-D Image Classification" @default.
• W3161926058 cites W1977295820 @default.
• W3161926058 cites W2009832130 @default.
• W3161926058 cites W2044535169 @default.
• W3161926058 cites W2073256350 @default.
• W3161926058 cites W2111619626 @default.
• W3161926058 cites W2117696986 @default.
• W3161926058 cites W2119112357 @default.
• W3161926058 cites W2257979135 @default.
• W3161926058 cites W2289252105 @default.
• W3161926058 cites W2554302513 @default.
• W3161926058 cites W2581082771 @default.
• W3161926058 cites W2593877424 @default.
• W3161926058 cites W2604319603 @default.
• W3161926058 cites W2621645821 @default.
• W3161926058 cites W2798170643 @default.
• W3161926058 cites W2804985843 @default.
• W3161926058 cites W2839844222 @default.
• W3161926058 cites W2919115771 @default.
• W3161926058 cites W2943925420 @default.
• W3161926058 cites W2963594949 @default.
• W3161926058 cites W3011442353 @default.
• W3161926058 cites W3021633728 @default.
• W3161926058 cites W3037927382 @default.
• W3161926058 cites W3044938306 @default.
• W3161926058 cites W3047017957 @default.
• W3161926058 cites W3139203094 @default.
• W3161926058 cites W3180876537 @default.
• W3161926058 cites W4244802472 @default.
• W3161926058 cites W4288083474 @default.
• W3161926058 doi "https://doi.org/10.1109/access.2022.3161738" @default.
• W3161926058 hasPublicationYear "2022" @default.
• W3161926058 type Work @default.
• W3161926058 sameAs 3161926058 @default.
• W3161926058 citedByCount "1" @default.
• W3161926058 countsByYear W31619260582023 @default.
• W3161926058 crossrefType "journal-article" @default.
• W3161926058 hasAuthorship W3161926058A5013814572 @default.
• W3161926058 hasAuthorship W3161926058A5020682177 @default.
• W3161926058 hasAuthorship W3161926058A5062311050 @default.
• W3161926058 hasAuthorship W3161926058A5078108496 @default.
• W3161926058 hasAuthorship W3161926058A5091151249 @default.
• W3161926058 hasBestOaLocation W31619260581 @default.
• W3161926058 hasConcept C11413529 @default.
• W3161926058 hasConcept C129844170 @default.
• W3161926058 hasConcept C154945302 @default.
• W3161926058 hasConcept C18903297 @default.
• W3161926058 hasConcept C190502265 @default.
• W3161926058 hasConcept C206588197 @default.
• W3161926058 hasConcept C2524010 @default.
• W3161926058 hasConcept C2776214188 @default.
• W3161926058 hasConcept C33923547 @default.
• W3161926058 hasConcept C41008148 @default.
• W3161926058 hasConcept C45374587 @default.
• W3161926058 hasConcept C50644808 @default.
• W3161926058 hasConcept C80444323 @default.
• W3161926058 hasConcept C86803240 @default.
• W3161926058 hasConceptScore W3161926058C11413529 @default.
• W3161926058 hasConceptScore W3161926058C129844170 @default.
• W3161926058 hasConceptScore W3161926058C154945302 @default.
• W3161926058 hasConceptScore W3161926058C18903297 @default.
• W3161926058 hasConceptScore W3161926058C190502265 @default.
• W3161926058 hasConceptScore W3161926058C206588197 @default.
• W3161926058 hasConceptScore W3161926058C2524010 @default.
• W3161926058 hasConceptScore W3161926058C2776214188 @default.
• W3161926058 hasConceptScore W3161926058C33923547 @default.
• W3161926058 hasConceptScore W3161926058C41008148 @default.
• W3161926058 hasConceptScore W3161926058C45374587 @default.
• W3161926058 hasConceptScore W3161926058C50644808 @default.
• W3161926058 hasConceptScore W3161926058C80444323 @default.
• W3161926058 hasConceptScore W3161926058C86803240 @default.
• W3161926058 hasFunder F4320322334 @default.
• W3161926058 hasLocation W31619260581 @default.
• W3161926058 hasLocation W31619260582 @default.
• W3161926058 hasLocation W31619260583 @default.
• W3161926058 hasOpenAccess W3161926058 @default.
• W3161926058 hasPrimaryLocation W31619260581 @default.
• W3161926058 hasRelatedWork W1853785581 @default.
• W3161926058 hasRelatedWork W2133796295 @default.
• W3161926058 hasRelatedWork W2559019371 @default.
• W3161926058 hasRelatedWork W2936783136 @default.
• W3161926058 hasRelatedWork W2947175736 @default.
• W3161926058 hasRelatedWork W2952959493 @default.
• W3161926058 hasRelatedWork W3037703370 @default.
• W3161926058 hasRelatedWork W4243928383 @default.
• W3161926058 hasRelatedWork W4283319738 @default.
• W3161926058 hasRelatedWork W4384112166 @default.
• W3161926058 hasVolume "10" @default.
• W3161926058 isParatext "false" @default.

• next