Matches in SemOpenAlex for { <https://semopenalex.org/work/W3014961572> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W3014961572 endingPage "62571" @default.
- W3014961572 startingPage "62561" @default.
- W3014961572 abstract "As a classic and well-performed deep convolutional neural network, DenseNet links every layer to each of its preceding layers via skip connections. However, the dense connectivity of the links leads to much redundance, consuming lots of computational resources. In this paper, to automatically prune redundant skip connections in DenseNet, we introduce a novel reinforcement learning method called automatic DenseNet sparsification (ADS). In ADS, we use adjacent matrix to represent dense connections in DenseNet, and design an agent using recurrent neural networks (RNNs) to sparsify the matrix, i. e. removing redundant skip connections in DenseNet. The validation accuracies of the sparsified DenseNets are used as rewards to update the agent, which promotes the agent to generate sparsified DenseNets with high performance. Extensive experiments demonstrate the effectiveness of ADS: The performance of the sparsified DenseNet surpasses not only the original DenseNet but related models; Moreover, the sparsified DenseNet has strong transferability when it is applied to new tasks. More importantly, ADS is very efficient. For the compression of a 40-layer DenseNet, it takes less than 1 day on a single GPU." @default.
- W3014961572 created "2020-04-10" @default.
- W3014961572 creator A5046728432 @default.
- W3014961572 creator A5051756041 @default.
- W3014961572 creator A5076230944 @default.
- W3014961572 creator A5085836254 @default.
- W3014961572 date "2020-01-01" @default.
- W3014961572 modified "2023-09-27" @default.
- W3014961572 title "Automatic DenseNet Sparsification" @default.
- W3014961572 cites W1996901117 @default.
- W3014961572 cites W2107878631 @default.
- W3014961572 cites W2119717200 @default.
- W3014961572 cites W2194775991 @default.
- W3014961572 cites W2331143823 @default.
- W3014961572 cites W2796265726 @default.
- W3014961572 cites W2886851211 @default.
- W3014961572 cites W2888270240 @default.
- W3014961572 cites W2916114993 @default.
- W3014961572 cites W2938366133 @default.
- W3014961572 cites W2941646209 @default.
- W3014961572 cites W2962851801 @default.
- W3014961572 cites W2963363373 @default.
- W3014961572 cites W2963446712 @default.
- W3014961572 cites W2963993763 @default.
- W3014961572 cites W2964081807 @default.
- W3014961572 cites W2964101377 @default.
- W3014961572 cites W2964212750 @default.
- W3014961572 cites W2964461714 @default.
- W3014961572 cites W2965658867 @default.
- W3014961572 cites W2967488183 @default.
- W3014961572 doi "https://doi.org/10.1109/access.2020.2984130" @default.
- W3014961572 hasPublicationYear "2020" @default.
- W3014961572 type Work @default.
- W3014961572 sameAs 3014961572 @default.
- W3014961572 citedByCount "12" @default.
- W3014961572 countsByYear W30149615722020 @default.
- W3014961572 countsByYear W30149615722021 @default.
- W3014961572 countsByYear W30149615722022 @default.
- W3014961572 countsByYear W30149615722023 @default.
- W3014961572 crossrefType "journal-article" @default.
- W3014961572 hasAuthorship W3014961572A5046728432 @default.
- W3014961572 hasAuthorship W3014961572A5051756041 @default.
- W3014961572 hasAuthorship W3014961572A5076230944 @default.
- W3014961572 hasAuthorship W3014961572A5085836254 @default.
- W3014961572 hasBestOaLocation W30149615721 @default.
- W3014961572 hasConcept C11413529 @default.
- W3014961572 hasConcept C119857082 @default.
- W3014961572 hasConcept C140331021 @default.
- W3014961572 hasConcept C154945302 @default.
- W3014961572 hasConcept C178790620 @default.
- W3014961572 hasConcept C185592680 @default.
- W3014961572 hasConcept C2779227376 @default.
- W3014961572 hasConcept C41008148 @default.
- W3014961572 hasConcept C61272859 @default.
- W3014961572 hasConcept C81363708 @default.
- W3014961572 hasConcept C97541855 @default.
- W3014961572 hasConceptScore W3014961572C11413529 @default.
- W3014961572 hasConceptScore W3014961572C119857082 @default.
- W3014961572 hasConceptScore W3014961572C140331021 @default.
- W3014961572 hasConceptScore W3014961572C154945302 @default.
- W3014961572 hasConceptScore W3014961572C178790620 @default.
- W3014961572 hasConceptScore W3014961572C185592680 @default.
- W3014961572 hasConceptScore W3014961572C2779227376 @default.
- W3014961572 hasConceptScore W3014961572C41008148 @default.
- W3014961572 hasConceptScore W3014961572C61272859 @default.
- W3014961572 hasConceptScore W3014961572C81363708 @default.
- W3014961572 hasConceptScore W3014961572C97541855 @default.
- W3014961572 hasFunder F4320321001 @default.
- W3014961572 hasFunder F4320335787 @default.
- W3014961572 hasLocation W30149615721 @default.
- W3014961572 hasOpenAccess W3014961572 @default.
- W3014961572 hasPrimaryLocation W30149615721 @default.
- W3014961572 hasRelatedWork W2923653485 @default.
- W3014961572 hasRelatedWork W2955938200 @default.
- W3014961572 hasRelatedWork W2957776456 @default.
- W3014961572 hasRelatedWork W2959276766 @default.
- W3014961572 hasRelatedWork W3005560120 @default.
- W3014961572 hasRelatedWork W3037422413 @default.
- W3014961572 hasRelatedWork W3209094908 @default.
- W3014961572 hasRelatedWork W4206669594 @default.
- W3014961572 hasRelatedWork W4210912933 @default.
- W3014961572 hasRelatedWork W4313050734 @default.
- W3014961572 hasVolume "8" @default.
- W3014961572 isParatext "false" @default.
- W3014961572 isRetracted "false" @default.
- W3014961572 magId "3014961572" @default.
- W3014961572 workType "article" @default.