Matches in SemOpenAlex for { <https://semopenalex.org/work/W3013326936> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W3013326936 abstract "Deep learning is an important field of computer research. Convolutional neural networks have received much attention in the field of image recognition because of their excellent recognition ability. The neural network contains a large number of parameters. For the training time-consuming problem of convolutional neural networks, how to effectively realize the computational acceleration in the training process is one of the hot topics in the field of deep learning research. In this paper, two parallel algorithms are designed based on single-thread block multi-thread and multi-thread block strategies. The former is a parallel algorithm between activation functions, and the latter is based on the former, and the parallelization algorithm inside the activation function. Finally, this paper designed a contrast experiment to verify the algorithm. The experimental results show that both algorithms have achieved a good acceleration effect, so that the GPU resources can be fully utilized to achieve the purpose of computational acceleration." @default.
- W3013326936 created "2020-04-03" @default.
- W3013326936 creator A5001139700 @default.
- W3013326936 creator A5001476875 @default.
- W3013326936 creator A5006879112 @default.
- W3013326936 date "2019-12-01" @default.
- W3013326936 modified "2023-10-06" @default.
- W3013326936 title "CUDA Optimization Method for Activation Function in Convolution Operation" @default.
- W3013326936 cites W1973112765 @default.
- W3013326936 cites W1982088757 @default.
- W3013326936 cites W1992118440 @default.
- W3013326936 cites W2084868624 @default.
- W3013326936 cites W2095231256 @default.
- W3013326936 cites W2118850909 @default.
- W3013326936 cites W2119348881 @default.
- W3013326936 cites W2163849015 @default.
- W3013326936 cites W2397037157 @default.
- W3013326936 cites W2588456389 @default.
- W3013326936 cites W2685231880 @default.
- W3013326936 cites W2766754718 @default.
- W3013326936 cites W2783066242 @default.
- W3013326936 cites W2807979944 @default.
- W3013326936 cites W3022086662 @default.
- W3013326936 doi "https://doi.org/10.1109/ispa-bdcloud-sustaincom-socialcom48970.2019.00079" @default.
- W3013326936 hasPublicationYear "2019" @default.
- W3013326936 type Work @default.
- W3013326936 sameAs 3013326936 @default.
- W3013326936 citedByCount "0" @default.
- W3013326936 crossrefType "proceedings-article" @default.
- W3013326936 hasAuthorship W3013326936A5001139700 @default.
- W3013326936 hasAuthorship W3013326936A5001476875 @default.
- W3013326936 hasAuthorship W3013326936A5006879112 @default.
- W3013326936 hasConcept C108583219 @default.
- W3013326936 hasConcept C111919701 @default.
- W3013326936 hasConcept C11413529 @default.
- W3013326936 hasConcept C117896860 @default.
- W3013326936 hasConcept C121332964 @default.
- W3013326936 hasConcept C138101251 @default.
- W3013326936 hasConcept C154945302 @default.
- W3013326936 hasConcept C173608175 @default.
- W3013326936 hasConcept C2524010 @default.
- W3013326936 hasConcept C2777210771 @default.
- W3013326936 hasConcept C2778119891 @default.
- W3013326936 hasConcept C33923547 @default.
- W3013326936 hasConcept C38365724 @default.
- W3013326936 hasConcept C41008148 @default.
- W3013326936 hasConcept C45347329 @default.
- W3013326936 hasConcept C50644808 @default.
- W3013326936 hasConcept C74650414 @default.
- W3013326936 hasConcept C81363708 @default.
- W3013326936 hasConceptScore W3013326936C108583219 @default.
- W3013326936 hasConceptScore W3013326936C111919701 @default.
- W3013326936 hasConceptScore W3013326936C11413529 @default.
- W3013326936 hasConceptScore W3013326936C117896860 @default.
- W3013326936 hasConceptScore W3013326936C121332964 @default.
- W3013326936 hasConceptScore W3013326936C138101251 @default.
- W3013326936 hasConceptScore W3013326936C154945302 @default.
- W3013326936 hasConceptScore W3013326936C173608175 @default.
- W3013326936 hasConceptScore W3013326936C2524010 @default.
- W3013326936 hasConceptScore W3013326936C2777210771 @default.
- W3013326936 hasConceptScore W3013326936C2778119891 @default.
- W3013326936 hasConceptScore W3013326936C33923547 @default.
- W3013326936 hasConceptScore W3013326936C38365724 @default.
- W3013326936 hasConceptScore W3013326936C41008148 @default.
- W3013326936 hasConceptScore W3013326936C45347329 @default.
- W3013326936 hasConceptScore W3013326936C50644808 @default.
- W3013326936 hasConceptScore W3013326936C74650414 @default.
- W3013326936 hasConceptScore W3013326936C81363708 @default.
- W3013326936 hasLocation W30133269361 @default.
- W3013326936 hasOpenAccess W3013326936 @default.
- W3013326936 hasPrimaryLocation W30133269361 @default.
- W3013326936 hasRelatedWork W12519965 @default.
- W3013326936 hasRelatedWork W131680 @default.
- W3013326936 hasRelatedWork W13678974 @default.
- W3013326936 hasRelatedWork W14085413 @default.
- W3013326936 hasRelatedWork W1889210 @default.
- W3013326936 hasRelatedWork W4136762 @default.
- W3013326936 hasRelatedWork W564427 @default.
- W3013326936 hasRelatedWork W758469 @default.
- W3013326936 hasRelatedWork W8203773 @default.
- W3013326936 hasRelatedWork W9190101 @default.
- W3013326936 isParatext "false" @default.
- W3013326936 isRetracted "false" @default.
- W3013326936 magId "3013326936" @default.
- W3013326936 workType "article" @default.