FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2935391837> ?p ?o ?g. }

• W2935391837 abstract "The redundancy is widely recognized in Convolutional Neural Networks (CNNs), which enables to remove unimportant filters from convolutional layers so as to slim the network with acceptable performance drop. Inspired by the linear and combinational properties of convolution, we seek to make some filters increasingly close and eventually identical for network slimming. To this end, we propose Centripetal SGD (C-SGD), a novel optimization method, which can train several filters to collapse into a single point in the parameter hyperspace. When the training is completed, the removal of the identical filters can trim the network with NO performance loss, thus no finetuning is needed. By doing so, we have partly solved an open problem of constrained filter pruning on CNNs with complicated structure, where some layers must be pruned following others. Our experimental results on CIFAR-10 and ImageNet have justified the effectiveness of C-SGD-based filter pruning. Moreover, we have provided empirical evidences for the assumption that the redundancy in deep neural networks helps the convergence of training by showing that a redundant CNN trained using C-SGD outperforms a normally trained counterpart with the equivalent width." @default.
• W2935391837 created "2019-04-11" @default.
• W2935391837 creator A5011301418 @default.
• W2935391837 creator A5035519525 @default.
• W2935391837 creator A5046605531 @default.
• W2935391837 creator A5057732142 @default.
• W2935391837 date "2019-04-08" @default.
• W2935391837 modified "2023-09-23" @default.
• W2935391837 title "Centripetal SGD for Pruning Very Deep Convolutional Networks with Complicated Structure" @default.
• W2935391837 cites W1532826273 @default.
• W2935391837 cites W1570197553 @default.
• W2935391837 cites W1686810756 @default.
• W2935391837 cites W1690739335 @default.
• W2935391837 cites W1789336918 @default.
• W2935391837 cites W1821462560 @default.
• W2935391837 cites W1836465849 @default.
• W2935391837 cites W1919191429 @default.
• W2935391837 cites W1922123711 @default.
• W2935391837 cites W1935978687 @default.
• W2935391837 cites W1984915212 @default.
• W2935391837 cites W2058641082 @default.
• W2935391837 cites W2072458630 @default.
• W2935391837 cites W2104636679 @default.
• W2935391837 cites W2108598243 @default.
• W2935391837 cites W2114766824 @default.
• W2935391837 cites W2117130368 @default.
• W2935391837 cites W2123469553 @default.
• W2935391837 cites W2125389748 @default.
• W2935391837 cites W2126333738 @default.
• W2935391837 cites W2134797427 @default.
• W2935391837 cites W2144354855 @default.
• W2935391837 cites W2144513243 @default.
• W2935391837 cites W2154579312 @default.
• W2935391837 cites W2163605009 @default.
• W2935391837 cites W2167215970 @default.
• W2935391837 cites W2182629226 @default.
• W2935391837 cites W2194775991 @default.
• W2935391837 cites W2233116163 @default.
• W2935391837 cites W2271840356 @default.
• W2935391837 cites W2276892413 @default.
• W2935391837 cites W2294543795 @default.
• W2935391837 cites W2300242332 @default.
• W2935391837 cites W2319920447 @default.
• W2935391837 cites W2495425901 @default.
• W2935391837 cites W2515385951 @default.
• W2935391837 cites W2520760693 @default.
• W2935391837 cites W2554242204 @default.
• W2935391837 cites W2554931888 @default.
• W2935391837 cites W2619444510 @default.
• W2935391837 cites W2707890836 @default.
• W2935391837 cites W2741232386 @default.
• W2935391837 cites W2785761580 @default.
• W2935391837 cites W2788715907 @default.
• W2935391837 cites W2807961551 @default.
• W2935391837 cites W2900493439 @default.
• W2935391837 cites W2901724498 @default.
• W2935391837 cites W2912648642 @default.
• W2935391837 cites W2949941638 @default.
• W2935391837 cites W2950967261 @default.
• W2935391837 cites W2951569836 @default.
• W2935391837 cites W2951977814 @default.
• W2935391837 cites W2952899695 @default.
• W2935391837 cites W2962851801 @default.
• W2935391837 cites W2962988160 @default.
• W2935391837 cites W2963000224 @default.
• W2935391837 cites W2963094099 @default.
• W2935391837 cites W2963145730 @default.
• W2935391837 cites W2963363373 @default.
• W2935391837 cites W2963374099 @default.
• W2935391837 cites W2963674932 @default.
• W2935391837 cites W2963981420 @default.
• W2935391837 cites W2964152344 @default.
• W2935391837 cites W2964233199 @default.
• W2935391837 cites W2964299589 @default.
• W2935391837 cites W3028304412 @default.
• W2935391837 cites W3118608800 @default.
• W2935391837 cites W3013573966 @default.
• W2935391837 doi "https://doi.org/10.48550/arxiv.1904.03837" @default.
• W2935391837 hasPublicationYear "2019" @default.
• W2935391837 type Work @default.
• W2935391837 sameAs 2935391837 @default.
• W2935391837 citedByCount "7" @default.
• W2935391837 countsByYear W29353918372019 @default.
• W2935391837 countsByYear W29353918372020 @default.
• W2935391837 crossrefType "posted-content" @default.
• W2935391837 hasAuthorship W2935391837A5011301418 @default.
• W2935391837 hasAuthorship W2935391837A5035519525 @default.
• W2935391837 hasAuthorship W2935391837A5046605531 @default.
• W2935391837 hasAuthorship W2935391837A5057732142 @default.
• W2935391837 hasBestOaLocation W29353918371 @default.
• W2935391837 hasConcept C106131492 @default.
• W2935391837 hasConcept C108010975 @default.
• W2935391837 hasConcept C111919701 @default.
• W2935391837 hasConcept C11413529 @default.
• W2935391837 hasConcept C119857082 @default.
• W2935391837 hasConcept C152124472 @default.
• W2935391837 hasConcept C154945302 @default.
• W2935391837 hasConcept C162324750 @default.
• W2935391837 hasConcept C2777303404 @default.
• W2935391837 hasConcept C2984842247 @default.

• next