Matches in SemOpenAlex for { <https://semopenalex.org/work/W2585683604> ?p ?o ?g. }
- W2585683604 abstract "Whereas CNNs have demonstrated immense progress in many vision problems, they suffer from a dependence on monumental amounts of labeled training data. On the other hand, dictionary learning does not scale to the size of problems that CNNs can handle, despite being very effective at low-level vision tasks such as denoising and inpainting. Recently, interest has grown in adapting dictionary learning methods for supervised tasks such as classification and inverse problems. We propose two new network layers that are based on dictionary learning: a sparse factorization layer and a convolutional sparse factorization layer, analogous to fully-connected and convolutional layers, respectively. Using our derivations, these layers can be dropped in to existing CNNs, trained together in an end-to-end fashion with back-propagation, and leverage semisupervision in ways classical CNNs cannot. We experimentally compare networks with these two new layers against a baseline CNN. Our results demonstrate that networks with either of the sparse factorization layers are able to outperform classical CNNs when supervised data are few. They also show performance improvements in certain tasks when compared to the CNN with no sparse factorization layers with the same exact number of parameters." @default.
- W2585683604 created "2017-02-10" @default.
- W2585683604 creator A5036212055 @default.
- W2585683604 creator A5079301873 @default.
- W2585683604 date "2016-12-14" @default.
- W2585683604 modified "2023-09-26" @default.
- W2585683604 title "Sparse Factorization Layers for Neural Networks with Limited Supervision" @default.
- W2585683604 cites W1576278180 @default.
- W2585683604 cites W1591116419 @default.
- W2585683604 cites W1686810756 @default.
- W2585683604 cites W1853900790 @default.
- W2585683604 cites W1861492603 @default.
- W2585683604 cites W1932198206 @default.
- W2585683604 cites W2004915807 @default.
- W2585683604 cites W2020719522 @default.
- W2585683604 cites W2027805700 @default.
- W2585683604 cites W2063978378 @default.
- W2585683604 cites W2066915075 @default.
- W2585683604 cites W2095705004 @default.
- W2585683604 cites W2100495367 @default.
- W2585683604 cites W2108598243 @default.
- W2585683604 cites W2108665656 @default.
- W2585683604 cites W2111719156 @default.
- W2585683604 cites W2112447569 @default.
- W2585683604 cites W2112796928 @default.
- W2585683604 cites W2114122776 @default.
- W2585683604 cites W2117259536 @default.
- W2585683604 cites W2117539524 @default.
- W2585683604 cites W2118688707 @default.
- W2585683604 cites W2122825543 @default.
- W2585683604 cites W2143104527 @default.
- W2585683604 cites W2155284717 @default.
- W2585683604 cites W2162033023 @default.
- W2585683604 cites W2163112044 @default.
- W2585683604 cites W2169488311 @default.
- W2585683604 cites W2194775991 @default.
- W2585683604 cites W2254140767 @default.
- W2585683604 cites W2524365899 @default.
- W2585683604 cites W2950064337 @default.
- W2585683604 cites W2951242004 @default.
- W2585683604 cites W2953066166 @default.
- W2585683604 cites W2963758027 @default.
- W2585683604 cites W3146803896 @default.
- W2585683604 cites W44815768 @default.
- W2585683604 cites W4919037 @default.
- W2585683604 hasPublicationYear "2016" @default.
- W2585683604 type Work @default.
- W2585683604 sameAs 2585683604 @default.
- W2585683604 citedByCount "1" @default.
- W2585683604 countsByYear W25856836042017 @default.
- W2585683604 crossrefType "posted-content" @default.
- W2585683604 hasAuthorship W2585683604A5036212055 @default.
- W2585683604 hasAuthorship W2585683604A5079301873 @default.
- W2585683604 hasConcept C108583219 @default.
- W2585683604 hasConcept C11413529 @default.
- W2585683604 hasConcept C115961682 @default.
- W2585683604 hasConcept C11727466 @default.
- W2585683604 hasConcept C119857082 @default.
- W2585683604 hasConcept C124066611 @default.
- W2585683604 hasConcept C153083717 @default.
- W2585683604 hasConcept C153180895 @default.
- W2585683604 hasConcept C154945302 @default.
- W2585683604 hasConcept C187834632 @default.
- W2585683604 hasConcept C2988886741 @default.
- W2585683604 hasConcept C41008148 @default.
- W2585683604 hasConcept C81363708 @default.
- W2585683604 hasConceptScore W2585683604C108583219 @default.
- W2585683604 hasConceptScore W2585683604C11413529 @default.
- W2585683604 hasConceptScore W2585683604C115961682 @default.
- W2585683604 hasConceptScore W2585683604C11727466 @default.
- W2585683604 hasConceptScore W2585683604C119857082 @default.
- W2585683604 hasConceptScore W2585683604C124066611 @default.
- W2585683604 hasConceptScore W2585683604C153083717 @default.
- W2585683604 hasConceptScore W2585683604C153180895 @default.
- W2585683604 hasConceptScore W2585683604C154945302 @default.
- W2585683604 hasConceptScore W2585683604C187834632 @default.
- W2585683604 hasConceptScore W2585683604C2988886741 @default.
- W2585683604 hasConceptScore W2585683604C41008148 @default.
- W2585683604 hasConceptScore W2585683604C81363708 @default.
- W2585683604 hasLocation W25856836041 @default.
- W2585683604 hasOpenAccess W2585683604 @default.
- W2585683604 hasPrimaryLocation W25856836041 @default.
- W2585683604 hasRelatedWork W189277179 @default.
- W2585683604 hasRelatedWork W2123045220 @default.
- W2585683604 hasRelatedWork W2205378640 @default.
- W2585683604 hasRelatedWork W2553894218 @default.
- W2585683604 hasRelatedWork W2599764753 @default.
- W2585683604 hasRelatedWork W2605177561 @default.
- W2585683604 hasRelatedWork W2799042817 @default.
- W2585683604 hasRelatedWork W2887286714 @default.
- W2585683604 hasRelatedWork W2891179104 @default.
- W2585683604 hasRelatedWork W2892839088 @default.
- W2585683604 hasRelatedWork W2901800056 @default.
- W2585683604 hasRelatedWork W2935959962 @default.
- W2585683604 hasRelatedWork W2966140116 @default.
- W2585683604 hasRelatedWork W2983108667 @default.
- W2585683604 hasRelatedWork W2983993072 @default.
- W2585683604 hasRelatedWork W3004718729 @default.
- W2585683604 hasRelatedWork W3098280677 @default.
- W2585683604 hasRelatedWork W3123338478 @default.