Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952164265> ?p ?o ?g. }
- W2952164265 abstract "It is not easy to design and run Convolutional Neural Networks (CNNs) due to: 1) finding the optimal number of filters (i.e., the width) at each layer is tricky, given an architecture; and 2) the computational intensity of CNNs impedes the deployment on computationally limited devices. Oracle Pruning is designed to remove the unimportant filters from a well-trained CNN, which estimates the filters' importance by ablating them in turn and evaluating the model, thus delivers high accuracy but suffers from intolerable time complexity, and requires a given resulting width but cannot automatically find it. To address these problems, we propose Approximated Oracle Filter Pruning (AOFP), which keeps searching for the least important filters in a binary search manner, makes pruning attempts by masking out filters randomly, accumulates the resulting errors, and finetunes the model via a multi-path framework. As AOFP enables simultaneous pruning on multiple layers, we can prune an existing very deep CNN with acceptable time cost, negligible accuracy drop, and no heuristic knowledge, or re-design a model which exerts higher accuracy and faster inference." @default.
- W2952164265 created "2019-06-27" @default.
- W2952164265 creator A5011301418 @default.
- W2952164265 creator A5035519525 @default.
- W2952164265 creator A5046605531 @default.
- W2952164265 creator A5054311881 @default.
- W2952164265 creator A5057732142 @default.
- W2952164265 date "2019-05-12" @default.
- W2952164265 modified "2023-09-27" @default.
- W2952164265 title "Approximated Oracle Filter Pruning for Destructive CNN Width Optimization" @default.
- W2952164265 cites W1538131130 @default.
- W2952164265 cites W1686810756 @default.
- W2952164265 cites W1821462560 @default.
- W2952164265 cites W1836465849 @default.
- W2952164265 cites W1984915212 @default.
- W2952164265 cites W2013825190 @default.
- W2952164265 cites W2095705004 @default.
- W2952164265 cites W2108598243 @default.
- W2952164265 cites W2114766824 @default.
- W2952164265 cites W2117130368 @default.
- W2952164265 cites W2123469553 @default.
- W2952164265 cites W2125389748 @default.
- W2952164265 cites W2126333738 @default.
- W2952164265 cites W2154579312 @default.
- W2952164265 cites W2194775991 @default.
- W2952164265 cites W2276892413 @default.
- W2952164265 cites W2495425901 @default.
- W2952164265 cites W2515385951 @default.
- W2952164265 cites W2554242204 @default.
- W2952164265 cites W2554931888 @default.
- W2952164265 cites W2572009792 @default.
- W2952164265 cites W2582745083 @default.
- W2952164265 cites W2619444510 @default.
- W2952164265 cites W2707890836 @default.
- W2952164265 cites W2741232386 @default.
- W2952164265 cites W2757143157 @default.
- W2952164265 cites W2785253905 @default.
- W2952164265 cites W2786054724 @default.
- W2952164265 cites W2788715907 @default.
- W2952164265 cites W2883111419 @default.
- W2952164265 cites W2897319353 @default.
- W2952164265 cites W2919192593 @default.
- W2952164265 cites W2935391837 @default.
- W2952164265 cites W2950967261 @default.
- W2952164265 cites W2962851801 @default.
- W2952164265 cites W2963000224 @default.
- W2952164265 cites W2963145730 @default.
- W2952164265 cites W2963363373 @default.
- W2952164265 cites W2963674932 @default.
- W2952164265 cites W2963981420 @default.
- W2952164265 cites W2964233199 @default.
- W2952164265 cites W2964299589 @default.
- W2952164265 cites W3028304412 @default.
- W2952164265 cites W3118608800 @default.
- W2952164265 cites W3013573966 @default.
- W2952164265 hasPublicationYear "2019" @default.
- W2952164265 type Work @default.
- W2952164265 sameAs 2952164265 @default.
- W2952164265 citedByCount "9" @default.
- W2952164265 countsByYear W29521642652019 @default.
- W2952164265 countsByYear W29521642652020 @default.
- W2952164265 countsByYear W29521642652021 @default.
- W2952164265 crossrefType "posted-content" @default.
- W2952164265 hasAuthorship W2952164265A5011301418 @default.
- W2952164265 hasAuthorship W2952164265A5035519525 @default.
- W2952164265 hasAuthorship W2952164265A5046605531 @default.
- W2952164265 hasAuthorship W2952164265A5054311881 @default.
- W2952164265 hasAuthorship W2952164265A5057732142 @default.
- W2952164265 hasConcept C106131492 @default.
- W2952164265 hasConcept C108010975 @default.
- W2952164265 hasConcept C11413529 @default.
- W2952164265 hasConcept C115903868 @default.
- W2952164265 hasConcept C153180895 @default.
- W2952164265 hasConcept C154945302 @default.
- W2952164265 hasConcept C173801870 @default.
- W2952164265 hasConcept C199360897 @default.
- W2952164265 hasConcept C2776214188 @default.
- W2952164265 hasConcept C2777735758 @default.
- W2952164265 hasConcept C31972630 @default.
- W2952164265 hasConcept C41008148 @default.
- W2952164265 hasConcept C55166926 @default.
- W2952164265 hasConcept C6557445 @default.
- W2952164265 hasConcept C81363708 @default.
- W2952164265 hasConcept C86803240 @default.
- W2952164265 hasConceptScore W2952164265C106131492 @default.
- W2952164265 hasConceptScore W2952164265C108010975 @default.
- W2952164265 hasConceptScore W2952164265C11413529 @default.
- W2952164265 hasConceptScore W2952164265C115903868 @default.
- W2952164265 hasConceptScore W2952164265C153180895 @default.
- W2952164265 hasConceptScore W2952164265C154945302 @default.
- W2952164265 hasConceptScore W2952164265C173801870 @default.
- W2952164265 hasConceptScore W2952164265C199360897 @default.
- W2952164265 hasConceptScore W2952164265C2776214188 @default.
- W2952164265 hasConceptScore W2952164265C2777735758 @default.
- W2952164265 hasConceptScore W2952164265C31972630 @default.
- W2952164265 hasConceptScore W2952164265C41008148 @default.
- W2952164265 hasConceptScore W2952164265C55166926 @default.
- W2952164265 hasConceptScore W2952164265C6557445 @default.
- W2952164265 hasConceptScore W2952164265C81363708 @default.
- W2952164265 hasConceptScore W2952164265C86803240 @default.