Matches in SemOpenAlex for { <https://semopenalex.org/work/W3212035326> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W3212035326 abstract "Adder neural networks (AdderNets) have shown impressive performance on image classification with only addition operations, which are more energy efficient than traditional convolutional neural networks built with multiplications. Compared with classification, there is a strong demand on reducing the energy consumption of modern object detectors via AdderNets for real-world applications such as autonomous driving and face detection. In this paper, we present an empirical study of AdderNets for object detection. We first reveal that the batch normalization statistics in the pre-trained adder backbone should not be frozen, since the relatively large feature variance of AdderNets. Moreover, we insert more shortcut connections in the neck part and design a new feature fusion architecture for avoiding the sparse features of adder layers. We present extensive ablation studies to explore several design choices of adder detectors. Comparisons with state-of-the-arts are conducted on COCO and PASCAL VOC benchmarks. Specifically, the proposed Adder FCOS achieves a 37.8% AP on the COCO val set, demonstrating comparable performance to that of the convolutional counterpart with an about $1.4times$ energy reduction." @default.
- W3212035326 created "2021-11-22" @default.
- W3212035326 creator A5001529504 @default.
- W3212035326 creator A5006817088 @default.
- W3212035326 creator A5062581018 @default.
- W3212035326 creator A5066602150 @default.
- W3212035326 creator A5084617283 @default.
- W3212035326 date "2021-12-27" @default.
- W3212035326 modified "2023-09-30" @default.
- W3212035326 title "An Empirical Study of Adder Neural Networks for Object Detection" @default.
- W3212035326 doi "https://doi.org/10.48550/arxiv.2112.13608" @default.
- W3212035326 hasPublicationYear "2021" @default.
- W3212035326 type Work @default.
- W3212035326 sameAs 3212035326 @default.
- W3212035326 citedByCount "0" @default.
- W3212035326 crossrefType "posted-content" @default.
- W3212035326 hasAuthorship W3212035326A5001529504 @default.
- W3212035326 hasAuthorship W3212035326A5006817088 @default.
- W3212035326 hasAuthorship W3212035326A5062581018 @default.
- W3212035326 hasAuthorship W3212035326A5066602150 @default.
- W3212035326 hasAuthorship W3212035326A5084617283 @default.
- W3212035326 hasBestOaLocation W32120353261 @default.
- W3212035326 hasConcept C113775141 @default.
- W3212035326 hasConcept C119599485 @default.
- W3212035326 hasConcept C127413603 @default.
- W3212035326 hasConcept C136886441 @default.
- W3212035326 hasConcept C144024400 @default.
- W3212035326 hasConcept C153180895 @default.
- W3212035326 hasConcept C154945302 @default.
- W3212035326 hasConcept C15744967 @default.
- W3212035326 hasConcept C164620267 @default.
- W3212035326 hasConcept C19165224 @default.
- W3212035326 hasConcept C199360897 @default.
- W3212035326 hasConcept C2742236 @default.
- W3212035326 hasConcept C2776151529 @default.
- W3212035326 hasConcept C2780767217 @default.
- W3212035326 hasConcept C41008148 @default.
- W3212035326 hasConcept C542102704 @default.
- W3212035326 hasConcept C75608658 @default.
- W3212035326 hasConcept C76155785 @default.
- W3212035326 hasConcept C81363708 @default.
- W3212035326 hasConcept C82876162 @default.
- W3212035326 hasConcept C94915269 @default.
- W3212035326 hasConceptScore W3212035326C113775141 @default.
- W3212035326 hasConceptScore W3212035326C119599485 @default.
- W3212035326 hasConceptScore W3212035326C127413603 @default.
- W3212035326 hasConceptScore W3212035326C136886441 @default.
- W3212035326 hasConceptScore W3212035326C144024400 @default.
- W3212035326 hasConceptScore W3212035326C153180895 @default.
- W3212035326 hasConceptScore W3212035326C154945302 @default.
- W3212035326 hasConceptScore W3212035326C15744967 @default.
- W3212035326 hasConceptScore W3212035326C164620267 @default.
- W3212035326 hasConceptScore W3212035326C19165224 @default.
- W3212035326 hasConceptScore W3212035326C199360897 @default.
- W3212035326 hasConceptScore W3212035326C2742236 @default.
- W3212035326 hasConceptScore W3212035326C2776151529 @default.
- W3212035326 hasConceptScore W3212035326C2780767217 @default.
- W3212035326 hasConceptScore W3212035326C41008148 @default.
- W3212035326 hasConceptScore W3212035326C542102704 @default.
- W3212035326 hasConceptScore W3212035326C75608658 @default.
- W3212035326 hasConceptScore W3212035326C76155785 @default.
- W3212035326 hasConceptScore W3212035326C81363708 @default.
- W3212035326 hasConceptScore W3212035326C82876162 @default.
- W3212035326 hasConceptScore W3212035326C94915269 @default.
- W3212035326 hasLocation W32120353261 @default.
- W3212035326 hasOpenAccess W3212035326 @default.
- W3212035326 hasPrimaryLocation W32120353261 @default.
- W3212035326 hasRelatedWork W2016839265 @default.
- W3212035326 hasRelatedWork W2424871898 @default.
- W3212035326 hasRelatedWork W2952466432 @default.
- W3212035326 hasRelatedWork W2995763445 @default.
- W3212035326 hasRelatedWork W3038672259 @default.
- W3212035326 hasRelatedWork W3100467467 @default.
- W3212035326 hasRelatedWork W3135818718 @default.
- W3212035326 hasRelatedWork W4297914674 @default.
- W3212035326 hasRelatedWork W4318827001 @default.
- W3212035326 hasRelatedWork W4385069170 @default.
- W3212035326 isParatext "false" @default.
- W3212035326 isRetracted "false" @default.
- W3212035326 magId "3212035326" @default.
- W3212035326 workType "article" @default.