Matches in SemOpenAlex for { <https://semopenalex.org/work/W3215673675> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W3215673675 endingPage "2047" @default.
- W3215673675 startingPage "2040" @default.
- W3215673675 abstract "Convolutional neural network (CNN) has a high recognition rate in image recognition and are used in embedded systems such as smartphones, robots and self-driving cars. Low-end FPGAs are candidates for embedded image recognition platforms because they achieve real-time performance at a low cost. However, CNN has significant parameters called weights and internal data called feature maps, which pose a challenge for FPGAs for performance and memory capacity. To solve these problems, we exploit a split-CNN and weight sparseness. The split-CNN reduces the memory footprint by splitting the feature map into smaller patches and allows the feature map to be stored in the FPGA's high-throughput on-chip memory. Weight sparseness reduces computational costs and achieves even higher performance. We designed a dedicated architecture of a sparse CNN and a memory buffering scheduling for a split-CNN and implemented this on the PYNQ-Z1 FPGA board with a low-end FPGA. An experiment on classification using VGG16 shows that our implementation is 3.1 times faster than the GPU, and 5.4 times faster than an existing FPGA implementation." @default.
- W3215673675 created "2021-12-06" @default.
- W3215673675 creator A5004085923 @default.
- W3215673675 creator A5016922897 @default.
- W3215673675 creator A5070734898 @default.
- W3215673675 date "2021-12-01" @default.
- W3215673675 modified "2023-10-03" @default.
- W3215673675 title "Weight Sparseness for a Feature-Map-Split-CNN Toward Low-Cost Embedded FPGAs" @default.
- W3215673675 cites W2094756095 @default.
- W3215673675 cites W2516141709 @default.
- W3215673675 cites W2559085405 @default.
- W3215673675 cites W2570343428 @default.
- W3215673675 cites W2613938013 @default.
- W3215673675 cites W2742152118 @default.
- W3215673675 cites W2794754997 @default.
- W3215673675 cites W2919115771 @default.
- W3215673675 cites W2931633209 @default.
- W3215673675 cites W2949619037 @default.
- W3215673675 cites W2952088389 @default.
- W3215673675 cites W2963881378 @default.
- W3215673675 cites W3034733718 @default.
- W3215673675 cites W3035532779 @default.
- W3215673675 doi "https://doi.org/10.1587/transinf.2021pap0011" @default.
- W3215673675 hasPublicationYear "2021" @default.
- W3215673675 type Work @default.
- W3215673675 sameAs 3215673675 @default.
- W3215673675 citedByCount "1" @default.
- W3215673675 countsByYear W32156736752023 @default.
- W3215673675 crossrefType "journal-article" @default.
- W3215673675 hasAuthorship W3215673675A5004085923 @default.
- W3215673675 hasAuthorship W3215673675A5016922897 @default.
- W3215673675 hasAuthorship W3215673675A5070734898 @default.
- W3215673675 hasBestOaLocation W32156736751 @default.
- W3215673675 hasConcept C111919701 @default.
- W3215673675 hasConcept C138885662 @default.
- W3215673675 hasConcept C149635348 @default.
- W3215673675 hasConcept C153180895 @default.
- W3215673675 hasConcept C154945302 @default.
- W3215673675 hasConcept C162324750 @default.
- W3215673675 hasConcept C173608175 @default.
- W3215673675 hasConcept C206729178 @default.
- W3215673675 hasConcept C21547014 @default.
- W3215673675 hasConcept C2776401178 @default.
- W3215673675 hasConcept C41008148 @default.
- W3215673675 hasConcept C41895202 @default.
- W3215673675 hasConcept C42935608 @default.
- W3215673675 hasConcept C74912251 @default.
- W3215673675 hasConcept C81363708 @default.
- W3215673675 hasConcept C9390403 @default.
- W3215673675 hasConceptScore W3215673675C111919701 @default.
- W3215673675 hasConceptScore W3215673675C138885662 @default.
- W3215673675 hasConceptScore W3215673675C149635348 @default.
- W3215673675 hasConceptScore W3215673675C153180895 @default.
- W3215673675 hasConceptScore W3215673675C154945302 @default.
- W3215673675 hasConceptScore W3215673675C162324750 @default.
- W3215673675 hasConceptScore W3215673675C173608175 @default.
- W3215673675 hasConceptScore W3215673675C206729178 @default.
- W3215673675 hasConceptScore W3215673675C21547014 @default.
- W3215673675 hasConceptScore W3215673675C2776401178 @default.
- W3215673675 hasConceptScore W3215673675C41008148 @default.
- W3215673675 hasConceptScore W3215673675C41895202 @default.
- W3215673675 hasConceptScore W3215673675C42935608 @default.
- W3215673675 hasConceptScore W3215673675C74912251 @default.
- W3215673675 hasConceptScore W3215673675C81363708 @default.
- W3215673675 hasConceptScore W3215673675C9390403 @default.
- W3215673675 hasIssue "12" @default.
- W3215673675 hasLocation W32156736751 @default.
- W3215673675 hasOpenAccess W3215673675 @default.
- W3215673675 hasPrimaryLocation W32156736751 @default.
- W3215673675 hasRelatedWork W2175746458 @default.
- W3215673675 hasRelatedWork W2732542196 @default.
- W3215673675 hasRelatedWork W2738221750 @default.
- W3215673675 hasRelatedWork W2760085659 @default.
- W3215673675 hasRelatedWork W2767651786 @default.
- W3215673675 hasRelatedWork W2883200793 @default.
- W3215673675 hasRelatedWork W2912288872 @default.
- W3215673675 hasRelatedWork W2940661641 @default.
- W3215673675 hasRelatedWork W3012978760 @default.
- W3215673675 hasRelatedWork W3093612317 @default.
- W3215673675 hasVolume "E104.D" @default.
- W3215673675 isParatext "false" @default.
- W3215673675 isRetracted "false" @default.
- W3215673675 magId "3215673675" @default.
- W3215673675 workType "article" @default.