Matches in SemOpenAlex for { <https://semopenalex.org/work/W3129155488> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W3129155488 abstract "To deploy powerful deep neural network (DNN) into smart, but resource limited IoT devices, many prior works have been proposed to compress DNN to reduce the network size and computation complexity with negligible accuracy degradation, such as weight quantization, network pruning, convolution decomposition, etc. However, by utilizing conventional DNN compression methods, a smaller, but fixed, network is generated from a relative large background model to achieve resource limited hardware acceleration. However, such optimization lacks the ability to adjust its structure in real-time to adapt for a dynamic computing hardware resource allocation and workloads. In this paper, we mainly review our two prior works [13, 15] to tackle this challenge, discussing how to construct a dynamic DNN by means of either uniform or non-uniform sub-nets generation methods. Moreover, to generate multiple nonuniform sub-nets, [15] needs to fully retrain the background model for each sub-net individually, named as multi-path method. To reduce the training cost, in this work, we further propose a single-path sub-nets generation method that can sample multiple sub-nets in different epochs within one training round. The constructed dynamic DNN, consisting of multiple sub-nets, provides the ability to run-time trade-off the inference accuracy and latency according to hardware resources and environment requirements. In the end, we study the the dynamic DNNs with different sub-nets generation methods on both CIFAR-10 and ImageNet dataset. We also present the run-time tuning of accuracy and latency on both GPU and CPU." @default.
- W3129155488 created "2021-02-15" @default.
- W3129155488 creator A5047916979 @default.
- W3129155488 creator A5063655609 @default.
- W3129155488 date "2021-01-18" @default.
- W3129155488 modified "2023-10-16" @default.
- W3129155488 title "Dynamic Neural Network to Enable Run-Time Trade-off between Accuracy and Latency" @default.
- W3129155488 cites W2108598243 @default.
- W3129155488 cites W2194775991 @default.
- W3129155488 cites W2963735024 @default.
- W3129155488 cites W2982644126 @default.
- W3129155488 cites W2998470761 @default.
- W3129155488 cites W3013202691 @default.
- W3129155488 doi "https://doi.org/10.1145/3394885.3431628" @default.
- W3129155488 hasPublicationYear "2021" @default.
- W3129155488 type Work @default.
- W3129155488 sameAs 3129155488 @default.
- W3129155488 citedByCount "0" @default.
- W3129155488 crossrefType "proceedings-article" @default.
- W3129155488 hasAuthorship W3129155488A5047916979 @default.
- W3129155488 hasAuthorship W3129155488A5063655609 @default.
- W3129155488 hasBestOaLocation W31291554881 @default.
- W3129155488 hasConcept C108010975 @default.
- W3129155488 hasConcept C113775141 @default.
- W3129155488 hasConcept C11413529 @default.
- W3129155488 hasConcept C120314980 @default.
- W3129155488 hasConcept C13164978 @default.
- W3129155488 hasConcept C149635348 @default.
- W3129155488 hasConcept C154945302 @default.
- W3129155488 hasConcept C173608175 @default.
- W3129155488 hasConcept C2776214188 @default.
- W3129155488 hasConcept C28855332 @default.
- W3129155488 hasConcept C41008148 @default.
- W3129155488 hasConcept C42935608 @default.
- W3129155488 hasConcept C50644808 @default.
- W3129155488 hasConcept C6557445 @default.
- W3129155488 hasConcept C76155785 @default.
- W3129155488 hasConcept C79403827 @default.
- W3129155488 hasConcept C82876162 @default.
- W3129155488 hasConcept C86803240 @default.
- W3129155488 hasConceptScore W3129155488C108010975 @default.
- W3129155488 hasConceptScore W3129155488C113775141 @default.
- W3129155488 hasConceptScore W3129155488C11413529 @default.
- W3129155488 hasConceptScore W3129155488C120314980 @default.
- W3129155488 hasConceptScore W3129155488C13164978 @default.
- W3129155488 hasConceptScore W3129155488C149635348 @default.
- W3129155488 hasConceptScore W3129155488C154945302 @default.
- W3129155488 hasConceptScore W3129155488C173608175 @default.
- W3129155488 hasConceptScore W3129155488C2776214188 @default.
- W3129155488 hasConceptScore W3129155488C28855332 @default.
- W3129155488 hasConceptScore W3129155488C41008148 @default.
- W3129155488 hasConceptScore W3129155488C42935608 @default.
- W3129155488 hasConceptScore W3129155488C50644808 @default.
- W3129155488 hasConceptScore W3129155488C6557445 @default.
- W3129155488 hasConceptScore W3129155488C76155785 @default.
- W3129155488 hasConceptScore W3129155488C79403827 @default.
- W3129155488 hasConceptScore W3129155488C82876162 @default.
- W3129155488 hasConceptScore W3129155488C86803240 @default.
- W3129155488 hasFunder F4320306076 @default.
- W3129155488 hasLocation W31291554881 @default.
- W3129155488 hasOpenAccess W3129155488 @default.
- W3129155488 hasPrimaryLocation W31291554881 @default.
- W3129155488 hasRelatedWork W1513409726 @default.
- W3129155488 hasRelatedWork W3088334494 @default.
- W3129155488 hasRelatedWork W3105802176 @default.
- W3129155488 hasRelatedWork W3188841390 @default.
- W3129155488 hasRelatedWork W4214908790 @default.
- W3129155488 hasRelatedWork W4286233754 @default.
- W3129155488 hasRelatedWork W4294982680 @default.
- W3129155488 hasRelatedWork W4318348488 @default.
- W3129155488 hasRelatedWork W4360764662 @default.
- W3129155488 hasRelatedWork W4362663447 @default.
- W3129155488 isParatext "false" @default.
- W3129155488 isRetracted "false" @default.
- W3129155488 magId "3129155488" @default.
- W3129155488 workType "article" @default.