Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313472034> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W4313472034 endingPage "192" @default.
- W4313472034 startingPage "192" @default.
- W4313472034 abstract "The programmability of FPGA suits the constantly changing convolutional neural network (CNN). However, several challenges arise when the previous FPGA-based accelerators update CNN. Firstly, although the model of RepVGG can balance accuracy and speed, it solely supports two types of kernels. Meanwhile, 8-bit integer-only quantization of PyTorch which can support various CNNs is seldom successfully supported by the FPGA-based accelerators. In addition, Winograd F(4 × 4, 3 × 3) uses less multiplication, but its transformation matrix contains irregular decimals, which could lead to accuracy problems. To tackle these issues, this paper proposes High-accuracy Branch-fused CNN Accelerator (HBCA): a toolchain and corresponding FPGA-based accelerator. The toolchain proposes inception-based branch–fused technique, which can support more types of kernels. Meanwhile, the accelerator proposes Winograd-quantization dual decimal–fuse techniques to balance accuracy and speed. In addition, this accelerator supports multi-types of kernels and proposes Winograd decomposed-part reuse, multi-mode BRAM & DSP and data reuse to increase power efficiency. Experiments show that HBCA is capable of supporting seven CNNs with different types of kernels and more branches. The accuracy loss is within 0.1% when compared to the quantized model. Furthermore, the power efficiency (GOPS/W) of Inception, ResNet and VGG is up to 226.6, 188.1 and 197.7, which are better than other FPGA-based CNN accelerators." @default.
- W4313472034 created "2023-01-06" @default.
- W4313472034 creator A5032764121 @default.
- W4313472034 creator A5040690539 @default.
- W4313472034 creator A5067710652 @default.
- W4313472034 creator A5081419061 @default.
- W4313472034 creator A5082509340 @default.
- W4313472034 date "2022-12-30" @default.
- W4313472034 modified "2023-09-23" @default.
- W4313472034 title "HBCA: A Toolchain for High-Accuracy Branch-Fused CNN Accelerator on FPGA with Dual-Decimal-Fused Technique" @default.
- W4313472034 cites W2172654076 @default.
- W4313472034 cites W2276486856 @default.
- W4313472034 cites W2294282016 @default.
- W4313472034 cites W2584311934 @default.
- W4313472034 cites W2604319603 @default.
- W4313472034 cites W2616014673 @default.
- W4313472034 cites W2785545076 @default.
- W4313472034 cites W2789246071 @default.
- W4313472034 cites W2896983500 @default.
- W4313472034 cites W2902895686 @default.
- W4313472034 cites W2913221350 @default.
- W4313472034 cites W2962814013 @default.
- W4313472034 cites W2963821229 @default.
- W4313472034 cites W2966011162 @default.
- W4313472034 cites W2969707169 @default.
- W4313472034 cites W2979455536 @default.
- W4313472034 cites W2981207549 @default.
- W4313472034 cites W3000160544 @default.
- W4313472034 cites W3008408165 @default.
- W4313472034 cites W3010819965 @default.
- W4313472034 cites W3025045095 @default.
- W4313472034 cites W3034429256 @default.
- W4313472034 cites W3167976421 @default.
- W4313472034 cites W3204714802 @default.
- W4313472034 cites W3217153855 @default.
- W4313472034 doi "https://doi.org/10.3390/electronics12010192" @default.
- W4313472034 hasPublicationYear "2022" @default.
- W4313472034 type Work @default.
- W4313472034 citedByCount "0" @default.
- W4313472034 crossrefType "journal-article" @default.
- W4313472034 hasAuthorship W4313472034A5032764121 @default.
- W4313472034 hasAuthorship W4313472034A5040690539 @default.
- W4313472034 hasAuthorship W4313472034A5067710652 @default.
- W4313472034 hasAuthorship W4313472034A5081419061 @default.
- W4313472034 hasAuthorship W4313472034A5082509340 @default.
- W4313472034 hasBestOaLocation W43134720341 @default.
- W4313472034 hasConcept C11413529 @default.
- W4313472034 hasConcept C13280743 @default.
- W4313472034 hasConcept C154945302 @default.
- W4313472034 hasConcept C173608175 @default.
- W4313472034 hasConcept C185798385 @default.
- W4313472034 hasConcept C199360897 @default.
- W4313472034 hasConcept C205649164 @default.
- W4313472034 hasConcept C2777062904 @default.
- W4313472034 hasConcept C2777904410 @default.
- W4313472034 hasConcept C28855332 @default.
- W4313472034 hasConcept C41008148 @default.
- W4313472034 hasConcept C42935608 @default.
- W4313472034 hasConcept C68339613 @default.
- W4313472034 hasConcept C81363708 @default.
- W4313472034 hasConcept C9390403 @default.
- W4313472034 hasConceptScore W4313472034C11413529 @default.
- W4313472034 hasConceptScore W4313472034C13280743 @default.
- W4313472034 hasConceptScore W4313472034C154945302 @default.
- W4313472034 hasConceptScore W4313472034C173608175 @default.
- W4313472034 hasConceptScore W4313472034C185798385 @default.
- W4313472034 hasConceptScore W4313472034C199360897 @default.
- W4313472034 hasConceptScore W4313472034C205649164 @default.
- W4313472034 hasConceptScore W4313472034C2777062904 @default.
- W4313472034 hasConceptScore W4313472034C2777904410 @default.
- W4313472034 hasConceptScore W4313472034C28855332 @default.
- W4313472034 hasConceptScore W4313472034C41008148 @default.
- W4313472034 hasConceptScore W4313472034C42935608 @default.
- W4313472034 hasConceptScore W4313472034C68339613 @default.
- W4313472034 hasConceptScore W4313472034C81363708 @default.
- W4313472034 hasConceptScore W4313472034C9390403 @default.
- W4313472034 hasIssue "1" @default.
- W4313472034 hasLocation W43134720341 @default.
- W4313472034 hasOpenAccess W4313472034 @default.
- W4313472034 hasPrimaryLocation W43134720341 @default.
- W4313472034 hasRelatedWork W1509211761 @default.
- W4313472034 hasRelatedWork W1800827217 @default.
- W4313472034 hasRelatedWork W1967627035 @default.
- W4313472034 hasRelatedWork W2053732522 @default.
- W4313472034 hasRelatedWork W2116172876 @default.
- W4313472034 hasRelatedWork W2319467001 @default.
- W4313472034 hasRelatedWork W2391299576 @default.
- W4313472034 hasRelatedWork W27867058 @default.
- W4313472034 hasRelatedWork W4294734199 @default.
- W4313472034 hasRelatedWork W4306156203 @default.
- W4313472034 hasVolume "12" @default.
- W4313472034 isParatext "false" @default.
- W4313472034 isRetracted "false" @default.
- W4313472034 workType "article" @default.