Matches in SemOpenAlex for { <https://semopenalex.org/work/W2554109421> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W2554109421 abstract "Recently published methods enable training of bitwise neural networks which allow reduced representation of down to a single bit per weight. We present a method that exploits ensemble decisions based on multiple stochastically sampled network models to increase performance figures of bitwise neural networks in terms of classification accuracy at inference. Our experiments with the CIFAR-10 and GTSRB datasets show that the performance of such network ensembles surpasses the performance of the high-precision base model. With this technique we achieve 5.81% best classification error on CIFAR-10 test set using bitwise networks. Concerning inference on embedded systems we evaluate these bitwise networks using a hardware efficient stochastic rounding procedure. Our work contributes to efficient embedded bitwise neural networks." @default.
- W2554109421 created "2016-11-30" @default.
- W2554109421 creator A5050903837 @default.
- W2554109421 creator A5056912495 @default.
- W2554109421 creator A5066899205 @default.
- W2554109421 creator A5090026131 @default.
- W2554109421 date "2016-11-20" @default.
- W2554109421 modified "2023-09-27" @default.
- W2554109421 title "Efficient Stochastic Inference of Bitwise Deep Neural Networks." @default.
- W2554109421 cites W1605005685 @default.
- W2554109421 cites W1841592590 @default.
- W2554109421 cites W1902934009 @default.
- W2554109421 cites W1970027940 @default.
- W2554109421 cites W2013305145 @default.
- W2554109421 cites W2035279479 @default.
- W2554109421 cites W2046542508 @default.
- W2554109421 cites W2154203817 @default.
- W2554109421 cites W2194775991 @default.
- W2554109421 cites W2279098554 @default.
- W2554109421 cites W2319920447 @default.
- W2554109421 cites W2949892913 @default.
- W2554109421 cites W2951978180 @default.
- W2554109421 cites W3118608800 @default.
- W2554109421 hasPublicationYear "2016" @default.
- W2554109421 type Work @default.
- W2554109421 sameAs 2554109421 @default.
- W2554109421 citedByCount "1" @default.
- W2554109421 countsByYear W25541094212019 @default.
- W2554109421 crossrefType "posted-content" @default.
- W2554109421 hasAuthorship W2554109421A5050903837 @default.
- W2554109421 hasAuthorship W2554109421A5056912495 @default.
- W2554109421 hasAuthorship W2554109421A5066899205 @default.
- W2554109421 hasAuthorship W2554109421A5090026131 @default.
- W2554109421 hasConcept C108583219 @default.
- W2554109421 hasConcept C119857082 @default.
- W2554109421 hasConcept C134765980 @default.
- W2554109421 hasConcept C154945302 @default.
- W2554109421 hasConcept C177264268 @default.
- W2554109421 hasConcept C199360897 @default.
- W2554109421 hasConcept C2776214188 @default.
- W2554109421 hasConcept C41008148 @default.
- W2554109421 hasConcept C50644808 @default.
- W2554109421 hasConceptScore W2554109421C108583219 @default.
- W2554109421 hasConceptScore W2554109421C119857082 @default.
- W2554109421 hasConceptScore W2554109421C134765980 @default.
- W2554109421 hasConceptScore W2554109421C154945302 @default.
- W2554109421 hasConceptScore W2554109421C177264268 @default.
- W2554109421 hasConceptScore W2554109421C199360897 @default.
- W2554109421 hasConceptScore W2554109421C2776214188 @default.
- W2554109421 hasConceptScore W2554109421C41008148 @default.
- W2554109421 hasConceptScore W2554109421C50644808 @default.
- W2554109421 hasLocation W25541094211 @default.
- W2554109421 hasOpenAccess W2554109421 @default.
- W2554109421 hasPrimaryLocation W25541094211 @default.
- W2554109421 hasRelatedWork W132113812 @default.
- W2554109421 hasRelatedWork W1964309505 @default.
- W2554109421 hasRelatedWork W1987653356 @default.
- W2554109421 hasRelatedWork W2117435895 @default.
- W2554109421 hasRelatedWork W2151441196 @default.
- W2554109421 hasRelatedWork W2464772092 @default.
- W2554109421 hasRelatedWork W2910090831 @default.
- W2554109421 hasRelatedWork W2914037680 @default.
- W2554109421 hasRelatedWork W2948059343 @default.
- W2554109421 hasRelatedWork W2963361348 @default.
- W2554109421 hasRelatedWork W2963946985 @default.
- W2554109421 hasRelatedWork W2969216006 @default.
- W2554109421 hasRelatedWork W2982234100 @default.
- W2554109421 hasRelatedWork W2984225221 @default.
- W2554109421 hasRelatedWork W2998574230 @default.
- W2554109421 hasRelatedWork W3012380922 @default.
- W2554109421 hasRelatedWork W3123962571 @default.
- W2554109421 hasRelatedWork W3126744695 @default.
- W2554109421 hasRelatedWork W3162829646 @default.
- W2554109421 hasRelatedWork W3199723613 @default.
- W2554109421 isParatext "false" @default.
- W2554109421 isRetracted "false" @default.
- W2554109421 magId "2554109421" @default.
- W2554109421 workType "article" @default.