Matches in SemOpenAlex for { <https://semopenalex.org/work/W4289550145> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W4289550145 endingPage "1876" @default.
- W4289550145 startingPage "1856" @default.
- W4289550145 abstract "We propose a stable, parallel approach to train Wasserstein conditional generative adversarial neural networks (W-CGANs) under the constraint of a fixed computational budget. Differently from previous distributed GANs training techniques, our approach avoids inter-process communications, reduces the risk of mode collapse and enhances scalability by using multiple generators, each one of them concurrently trained on a single data label. The use of the Wasserstein metric also reduces the risk of cycling by stabilizing the training of each generator. We illustrate the approach on the CIFAR10, CIFAR100, and ImageNet1k datasets, three standard benchmark image datasets, maintaining the original resolution of the images for each dataset. Performance is assessed in terms of scalability and final accuracy within a limited fixed computational time and computational resources. To measure accuracy, we use the inception score, the Fréchet inception distance, and image quality. An improvement in inception score and Fréchet inception distance is shown in comparison to previous results obtained by performing the parallel approach on deep convolutional conditional generative adversarial neural networks as well as an improvement of image quality of the new images created by the GANs approach. Weak scaling is attained on both datasets using up to 2000 NVIDIA V100 GPUs on the OLCF supercomputer Summit." @default.
- W4289550145 created "2022-08-03" @default.
- W4289550145 creator A5019108731 @default.
- W4289550145 creator A5051882926 @default.
- W4289550145 date "2022-08-03" @default.
- W4289550145 modified "2023-09-30" @default.
- W4289550145 title "Stable parallel training of Wasserstein conditional generative adversarial neural networks" @default.
- W4289550145 cites W1965555277 @default.
- W4289550145 cites W2146950091 @default.
- W4289550145 cites W2737777148 @default.
- W4289550145 cites W2899250169 @default.
- W4289550145 cites W2963017889 @default.
- W4289550145 cites W2973349625 @default.
- W4289550145 cites W3119186746 @default.
- W4289550145 cites W3133167265 @default.
- W4289550145 cites W4211108990 @default.
- W4289550145 cites W4287897619 @default.
- W4289550145 cites W4293775665 @default.
- W4289550145 doi "https://doi.org/10.1007/s11227-022-04721-y" @default.
- W4289550145 hasPublicationYear "2022" @default.
- W4289550145 type Work @default.
- W4289550145 citedByCount "2" @default.
- W4289550145 countsByYear W42895501452023 @default.
- W4289550145 crossrefType "journal-article" @default.
- W4289550145 hasAuthorship W4289550145A5019108731 @default.
- W4289550145 hasAuthorship W4289550145A5051882926 @default.
- W4289550145 hasBestOaLocation W42895501453 @default.
- W4289550145 hasConcept C115961682 @default.
- W4289550145 hasConcept C119857082 @default.
- W4289550145 hasConcept C121332964 @default.
- W4289550145 hasConcept C13280743 @default.
- W4289550145 hasConcept C153180895 @default.
- W4289550145 hasConcept C154945302 @default.
- W4289550145 hasConcept C162324750 @default.
- W4289550145 hasConcept C163258240 @default.
- W4289550145 hasConcept C176217482 @default.
- W4289550145 hasConcept C185798385 @default.
- W4289550145 hasConcept C205649164 @default.
- W4289550145 hasConcept C21547014 @default.
- W4289550145 hasConcept C2780992000 @default.
- W4289550145 hasConcept C41008148 @default.
- W4289550145 hasConcept C48044578 @default.
- W4289550145 hasConcept C50644808 @default.
- W4289550145 hasConcept C62520636 @default.
- W4289550145 hasConcept C77088390 @default.
- W4289550145 hasConcept C81363708 @default.
- W4289550145 hasConceptScore W4289550145C115961682 @default.
- W4289550145 hasConceptScore W4289550145C119857082 @default.
- W4289550145 hasConceptScore W4289550145C121332964 @default.
- W4289550145 hasConceptScore W4289550145C13280743 @default.
- W4289550145 hasConceptScore W4289550145C153180895 @default.
- W4289550145 hasConceptScore W4289550145C154945302 @default.
- W4289550145 hasConceptScore W4289550145C162324750 @default.
- W4289550145 hasConceptScore W4289550145C163258240 @default.
- W4289550145 hasConceptScore W4289550145C176217482 @default.
- W4289550145 hasConceptScore W4289550145C185798385 @default.
- W4289550145 hasConceptScore W4289550145C205649164 @default.
- W4289550145 hasConceptScore W4289550145C21547014 @default.
- W4289550145 hasConceptScore W4289550145C2780992000 @default.
- W4289550145 hasConceptScore W4289550145C41008148 @default.
- W4289550145 hasConceptScore W4289550145C48044578 @default.
- W4289550145 hasConceptScore W4289550145C50644808 @default.
- W4289550145 hasConceptScore W4289550145C62520636 @default.
- W4289550145 hasConceptScore W4289550145C77088390 @default.
- W4289550145 hasConceptScore W4289550145C81363708 @default.
- W4289550145 hasFunder F4320332359 @default.
- W4289550145 hasIssue "2" @default.
- W4289550145 hasLocation W42895501451 @default.
- W4289550145 hasLocation W42895501452 @default.
- W4289550145 hasLocation W42895501453 @default.
- W4289550145 hasOpenAccess W4289550145 @default.
- W4289550145 hasPrimaryLocation W42895501451 @default.
- W4289550145 hasRelatedWork W112744582 @default.
- W4289550145 hasRelatedWork W1992807924 @default.
- W4289550145 hasRelatedWork W2175746458 @default.
- W4289550145 hasRelatedWork W2789601449 @default.
- W4289550145 hasRelatedWork W2949389737 @default.
- W4289550145 hasRelatedWork W2963556241 @default.
- W4289550145 hasRelatedWork W3027997911 @default.
- W4289550145 hasRelatedWork W3093612317 @default.
- W4289550145 hasRelatedWork W3109429904 @default.
- W4289550145 hasRelatedWork W4287776258 @default.
- W4289550145 hasVolume "79" @default.
- W4289550145 isParatext "false" @default.
- W4289550145 isRetracted "false" @default.
- W4289550145 workType "article" @default.