FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4308694752> ?p ?o ?g. }

• W4308694752 endingPage "11414" @default.
• W4308694752 startingPage "11414" @default.
• W4308694752 abstract "Deep neural networks (DNNs) have achieved great success in the field of computer vision. The high requirements for memory and storage by DNNs make it difficult to apply them to mobile or embedded devices. Therefore, compression and structure optimization of deep neural networks have become a hot research topic. To eliminate redundant structures in deep convolutional neural networks (DCNNs), we propose an efficient filter pruning framework via deep reinforcement learning (DRL). The proposed framework is based on a deep deterministic policy gradient (DDPG) algorithm for filter pruning rate optimization. The main features of the proposed framework are as follows: (1) AA tailored reward function considering both accuracy and complexity of DCNN is proposed for the training of DDPG and (2) a novel filter sorting criterion based on Taylor expansion is developed for filter pruning selection. To illustrate the effectiveness of the proposed framework, extensive comparative studies on large public datasets and well-recognized DCNNs are conducted. The experimental results demonstrate that the Taylor-expansion-based filter sorting criterion is much better than the widely used minimum-weight-based criterion. More importantly, the proposed filter pruning framework can achieve over 10× parameter compression and 3× floating point operations (FLOPs) reduction while maintaining similar accuracy to the original network. The performance of the proposed framework is promising compared with state-of-the-art DRL-based filter pruning methods." @default.
• W4308694752 created "2022-11-14" @default.
• W4308694752 creator A5006594763 @default.
• W4308694752 creator A5067027007 @default.
• W4308694752 creator A5070952629 @default.
• W4308694752 creator A5080932283 @default.
• W4308694752 creator A5081704135 @default.
• W4308694752 date "2022-11-10" @default.
• W4308694752 modified "2023-10-14" @default.
• W4308694752 title "A Novel Filter-Level Deep Convolutional Neural Network Pruning Method Based on Deep Reinforcement Learning" @default.
• W4308694752 cites W2097117768 @default.
• W4308694752 cites W2145339207 @default.
• W4308694752 cites W2194775991 @default.
• W4308694752 cites W2554302513 @default.
• W4308694752 cites W2618530766 @default.
• W4308694752 cites W2808168148 @default.
• W4308694752 cites W2886851211 @default.
• W4308694752 cites W2894581376 @default.
• W4308694752 cites W2953308748 @default.
• W4308694752 cites W2962851801 @default.
• W4308694752 cites W2963363373 @default.
• W4308694752 cites W2963946985 @default.
• W4308694752 cites W2984618279 @default.
• W4308694752 cites W3034251466 @default.
• W4308694752 cites W3034307881 @default.
• W4308694752 cites W3090190726 @default.
• W4308694752 cites W3103320289 @default.
• W4308694752 cites W3103559770 @default.
• W4308694752 cites W3131107792 @default.
• W4308694752 cites W3177126286 @default.
• W4308694752 cites W4206719179 @default.
• W4308694752 cites W4225905866 @default.
• W4308694752 doi "https://doi.org/10.3390/app122211414" @default.
• W4308694752 hasPublicationYear "2022" @default.
• W4308694752 type Work @default.
• W4308694752 citedByCount "1" @default.
• W4308694752 countsByYear W43086947522023 @default.
• W4308694752 crossrefType "journal-article" @default.
• W4308694752 hasAuthorship W4308694752A5006594763 @default.
• W4308694752 hasAuthorship W4308694752A5067027007 @default.
• W4308694752 hasAuthorship W4308694752A5070952629 @default.
• W4308694752 hasAuthorship W4308694752A5080932283 @default.
• W4308694752 hasAuthorship W4308694752A5081704135 @default.
• W4308694752 hasBestOaLocation W43086947521 @default.
• W4308694752 hasConcept C106131492 @default.
• W4308694752 hasConcept C108010975 @default.
• W4308694752 hasConcept C108583219 @default.
• W4308694752 hasConcept C111696304 @default.
• W4308694752 hasConcept C11413529 @default.
• W4308694752 hasConcept C119857082 @default.
• W4308694752 hasConcept C13107197 @default.
• W4308694752 hasConcept C154945302 @default.
• W4308694752 hasConcept C173608175 @default.
• W4308694752 hasConcept C22597639 @default.
• W4308694752 hasConcept C31972630 @default.
• W4308694752 hasConcept C3826847 @default.
• W4308694752 hasConcept C41008148 @default.
• W4308694752 hasConcept C50644808 @default.
• W4308694752 hasConcept C6557445 @default.
• W4308694752 hasConcept C81363708 @default.
• W4308694752 hasConcept C86803240 @default.
• W4308694752 hasConcept C97541855 @default.
• W4308694752 hasConceptScore W4308694752C106131492 @default.
• W4308694752 hasConceptScore W4308694752C108010975 @default.
• W4308694752 hasConceptScore W4308694752C108583219 @default.
• W4308694752 hasConceptScore W4308694752C111696304 @default.
• W4308694752 hasConceptScore W4308694752C11413529 @default.
• W4308694752 hasConceptScore W4308694752C119857082 @default.
• W4308694752 hasConceptScore W4308694752C13107197 @default.
• W4308694752 hasConceptScore W4308694752C154945302 @default.
• W4308694752 hasConceptScore W4308694752C173608175 @default.
• W4308694752 hasConceptScore W4308694752C22597639 @default.
• W4308694752 hasConceptScore W4308694752C31972630 @default.
• W4308694752 hasConceptScore W4308694752C3826847 @default.
• W4308694752 hasConceptScore W4308694752C41008148 @default.
• W4308694752 hasConceptScore W4308694752C50644808 @default.
• W4308694752 hasConceptScore W4308694752C6557445 @default.
• W4308694752 hasConceptScore W4308694752C81363708 @default.
• W4308694752 hasConceptScore W4308694752C86803240 @default.
• W4308694752 hasConceptScore W4308694752C97541855 @default.
• W4308694752 hasFunder F4320321001 @default.
• W4308694752 hasIssue "22" @default.
• W4308694752 hasLocation W43086947521 @default.
• W4308694752 hasLocation W43086947522 @default.
• W4308694752 hasOpenAccess W4308694752 @default.
• W4308694752 hasPrimaryLocation W43086947521 @default.
• W4308694752 hasRelatedWork W2731899572 @default.
• W4308694752 hasRelatedWork W2898360562 @default.
• W4308694752 hasRelatedWork W2999805992 @default.
• W4308694752 hasRelatedWork W3116150086 @default.
• W4308694752 hasRelatedWork W3133861977 @default.
• W4308694752 hasRelatedWork W3183066147 @default.
• W4308694752 hasRelatedWork W3191582637 @default.
• W4308694752 hasRelatedWork W4200173597 @default.
• W4308694752 hasRelatedWork W4312417841 @default.
• W4308694752 hasRelatedWork W4321369474 @default.
• W4308694752 hasVolume "12" @default.
• W4308694752 isParatext "false" @default.

• next