FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3127492156> ?p ?o ?g. }

• W3127492156 abstract "Searching for network width is an effective way to slim deep neural networks with hardware budgets. With this aim, a one-shot supernet is usually leveraged as a performance evaluator to rank the performance wrt~different width. Nevertheless, current methods mainly follow a manually fixed weight sharing pattern, which is limited to distinguish the performance gap of different width. In this paper, to better evaluate each width, we propose a locally free weight sharing strategy (CafeNet) accordingly. In CafeNet, weights are more freely shared, and each width is jointly indicated by its base channels and free channels, where free channels are supposed to loCAte FrEely in a local zone to better represent each width. Besides, we propose to further reduce the search space by leveraging our introduced FLOPs-sensitive bins. As a result, our CafeNet can be trained stochastically and get optimized within a min-min strategy. Extensive experiments on ImageNet, CIFAR-10, CelebA and MS COCO dataset have verified our superiority comparing to other state-of-the-art baselines. For example, our method can further boost the benchmark NAS network EfficientNet-B0 by 0.41% via searching its width more delicately." @default.
• W3127492156 created "2021-02-15" @default.
• W3127492156 creator A5001529504 @default.
• W3127492156 creator A5011334334 @default.
• W3127492156 creator A5028771381 @default.
• W3127492156 creator A5065063835 @default.
• W3127492156 creator A5082630089 @default.
• W3127492156 creator A5083588442 @default.
• W3127492156 creator A5086529957 @default.
• W3127492156 date "2021-02-10" @default.
• W3127492156 modified "2023-09-27" @default.
• W3127492156 title "Locally Free Weight Sharing for Network Width Search" @default.
• W3127492156 cites W1548969695 @default.
• W3127492156 cites W1686810756 @default.
• W3127492156 cites W1821462560 @default.
• W3127492156 cites W1834627138 @default.
• W3127492156 cites W1971259134 @default.
• W3127492156 cites W2031489346 @default.
• W3127492156 cites W2108598243 @default.
• W3127492156 cites W2194775991 @default.
• W3127492156 cites W2565639579 @default.
• W3127492156 cites W2613718673 @default.
• W3127492156 cites W2743289088 @default.
• W3127492156 cites W2743473392 @default.
• W3127492156 cites W2810075754 @default.
• W3127492156 cites W2886851211 @default.
• W3127492156 cites W2896006880 @default.
• W3127492156 cites W2896409484 @default.
• W3127492156 cites W2928560789 @default.
• W3127492156 cites W2940478336 @default.
• W3127492156 cites W2945335799 @default.
• W3127492156 cites W2947681860 @default.
• W3127492156 cites W2949117887 @default.
• W3127492156 cites W2950014519 @default.
• W3127492156 cites W2950673314 @default.
• W3127492156 cites W2951868790 @default.
• W3127492156 cites W2951977814 @default.
• W3127492156 cites W2952164265 @default.
• W3127492156 cites W2952838799 @default.
• W3127492156 cites W2955425717 @default.
• W3127492156 cites W2959934193 @default.
• W3127492156 cites W2962851801 @default.
• W3127492156 cites W2962965870 @default.
• W3127492156 cites W2963094099 @default.
• W3127492156 cites W2963163009 @default.
• W3127492156 cites W2963363373 @default.
• W3127492156 cites W2964299589 @default.
• W3127492156 cites W2970604398 @default.
• W3127492156 cites W2970632972 @default.
• W3127492156 cites W2970790493 @default.
• W3127492156 cites W2984618279 @default.
• W3127492156 cites W2994749257 @default.
• W3127492156 cites W2996319087 @default.
• W3127492156 cites W2998310140 @default.
• W3127492156 cites W2998583092 @default.
• W3127492156 cites W3009043942 @default.
• W3127492156 cites W3011514561 @default.
• W3127492156 cites W3022018744 @default.
• W3127492156 cites W3028304412 @default.
• W3127492156 cites W3035189477 @default.
• W3127492156 cites W3035422913 @default.
• W3127492156 cites W3048954095 @default.
• W3127492156 cites W3093155142 @default.
• W3127492156 cites W3097702032 @default.
• W3127492156 cites W3102846346 @default.
• W3127492156 cites W3124794912 @default.
• W3127492156 cites W3099225546 @default.
• W3127492156 hasPublicationYear "2021" @default.
• W3127492156 type Work @default.
• W3127492156 sameAs 3127492156 @default.
• W3127492156 citedByCount "6" @default.
• W3127492156 countsByYear W31274921562021 @default.
• W3127492156 crossrefType "posted-content" @default.
• W3127492156 hasAuthorship W3127492156A5001529504 @default.
• W3127492156 hasAuthorship W3127492156A5011334334 @default.
• W3127492156 hasAuthorship W3127492156A5028771381 @default.
• W3127492156 hasAuthorship W3127492156A5065063835 @default.
• W3127492156 hasAuthorship W3127492156A5082630089 @default.
• W3127492156 hasAuthorship W3127492156A5083588442 @default.
• W3127492156 hasAuthorship W3127492156A5086529957 @default.
• W3127492156 hasConcept C11413529 @default.
• W3127492156 hasConcept C114614502 @default.
• W3127492156 hasConcept C13280743 @default.
• W3127492156 hasConcept C134306372 @default.
• W3127492156 hasConcept C154945302 @default.
• W3127492156 hasConcept C164226766 @default.
• W3127492156 hasConcept C173608175 @default.
• W3127492156 hasConcept C185798385 @default.
• W3127492156 hasConcept C205649164 @default.
• W3127492156 hasConcept C33923547 @default.
• W3127492156 hasConcept C3826847 @default.
• W3127492156 hasConcept C41008148 @default.
• W3127492156 hasConcept C42058472 @default.
• W3127492156 hasConcept C50644808 @default.
• W3127492156 hasConceptScore W3127492156C11413529 @default.
• W3127492156 hasConceptScore W3127492156C114614502 @default.
• W3127492156 hasConceptScore W3127492156C13280743 @default.
• W3127492156 hasConceptScore W3127492156C134306372 @default.
• W3127492156 hasConceptScore W3127492156C154945302 @default.
• W3127492156 hasConceptScore W3127492156C164226766 @default.

• next