FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2565257220> ?p ?o ?g. }

• W2565257220 endingPage "1214" @default.
• W2565257220 startingPage "1200" @default.
• W2565257220 abstract "Adoption of deep learning in image steganalysis is still in its initial stage. In this paper we propose a generic hybrid deep-learning framework for JPEG steganalysis incorporating the domain knowledge behind rich steganalytic models. Our proposed framework involves two main stages. The first stage is hand-crafted, corresponding to the convolution phase and the quantization & truncation phase of the rich models. The second stage is a compound deep neural network containing multiple deep subnets in which the model parameters are learned in the training procedure. We provided experimental evidences and theoretical reflections to argue that the introduction of threshold quantizers, though disable the gradient-descent-based learning of the bottom convolution phase, is indeed cost-effective. We have conducted extensive experiments on a large-scale dataset extracted from ImageNet. The primary dataset used in our experiments contains 500,000 cover images, while our largest dataset contains five million cover images. Our experiments show that the integration of quantization and truncation into deep-learning steganalyzers do boost the detection performance by a clear margin. Furthermore, we demonstrate that our framework is insensitive to JPEG blocking artifact alterations, and the learned model can be easily transferred to a different attacking target and even a different dataset. These properties are of critical importance in practical applications." @default.
• W2565257220 created "2017-01-06" @default.
• W2565257220 creator A5023503629 @default.
• W2565257220 creator A5036117084 @default.
• W2565257220 creator A5047964483 @default.
• W2565257220 creator A5061167775 @default.
• W2565257220 date "2018-05-01" @default.
• W2565257220 modified "2023-10-15" @default.
• W2565257220 title "Large-Scale JPEG Image Steganalysis Using Hybrid Deep-Learning Framework" @default.
• W2565257220 cites W1604614341 @default.
• W2565257220 cites W1663899164 @default.
• W2565257220 cites W1955882805 @default.
• W2565257220 cites W1965064373 @default.
• W2565257220 cites W1973466159 @default.
• W2565257220 cites W1976570511 @default.
• W2565257220 cites W2009130368 @default.
• W2565257220 cites W2055625960 @default.
• W2565257220 cites W2076063813 @default.
• W2565257220 cites W2081564928 @default.
• W2565257220 cites W2096156898 @default.
• W2565257220 cites W2097117768 @default.
• W2565257220 cites W2104705565 @default.
• W2565257220 cites W2124664712 @default.
• W2565257220 cites W2170598445 @default.
• W2565257220 cites W2192227561 @default.
• W2565257220 cites W2277839806 @default.
• W2565257220 cites W2322622188 @default.
• W2565257220 cites W2339370745 @default.
• W2565257220 cites W2412509443 @default.
• W2565257220 cites W2416075718 @default.
• W2565257220 cites W2514127746 @default.
• W2565257220 cites W2538511122 @default.
• W2565257220 cites W2543123829 @default.
• W2565257220 cites W2610979719 @default.
• W2565257220 cites W2735904389 @default.
• W2565257220 cites W2888845754 @default.
• W2565257220 cites W2963682422 @default.
• W2565257220 cites W2972355823 @default.
• W2565257220 cites W4301488149 @default.
• W2565257220 doi "https://doi.org/10.1109/tifs.2017.2779446" @default.
• W2565257220 hasPublicationYear "2018" @default.
• W2565257220 type Work @default.
• W2565257220 sameAs 2565257220 @default.
• W2565257220 citedByCount "140" @default.
• W2565257220 countsByYear W25652572202018 @default.
• W2565257220 countsByYear W25652572202019 @default.
• W2565257220 countsByYear W25652572202020 @default.
• W2565257220 countsByYear W25652572202021 @default.
• W2565257220 countsByYear W25652572202022 @default.
• W2565257220 countsByYear W25652572202023 @default.
• W2565257220 crossrefType "journal-article" @default.
• W2565257220 hasAuthorship W2565257220A5023503629 @default.
• W2565257220 hasAuthorship W2565257220A5036117084 @default.
• W2565257220 hasAuthorship W2565257220A5047964483 @default.
• W2565257220 hasAuthorship W2565257220A5061167775 @default.
• W2565257220 hasBestOaLocation W25652572202 @default.
• W2565257220 hasConcept C107368093 @default.
• W2565257220 hasConcept C108583219 @default.
• W2565257220 hasConcept C108801101 @default.
• W2565257220 hasConcept C11413529 @default.
• W2565257220 hasConcept C115961682 @default.
• W2565257220 hasConcept C119857082 @default.
• W2565257220 hasConcept C153180895 @default.
• W2565257220 hasConcept C154945302 @default.
• W2565257220 hasConcept C169805256 @default.
• W2565257220 hasConcept C198751489 @default.
• W2565257220 hasConcept C2221639 @default.
• W2565257220 hasConcept C28855332 @default.
• W2565257220 hasConcept C41008148 @default.
• W2565257220 hasConcept C45347329 @default.
• W2565257220 hasConcept C50644808 @default.
• W2565257220 hasConcept C81363708 @default.
• W2565257220 hasConceptScore W2565257220C107368093 @default.
• W2565257220 hasConceptScore W2565257220C108583219 @default.
• W2565257220 hasConceptScore W2565257220C108801101 @default.
• W2565257220 hasConceptScore W2565257220C11413529 @default.
• W2565257220 hasConceptScore W2565257220C115961682 @default.
• W2565257220 hasConceptScore W2565257220C119857082 @default.
• W2565257220 hasConceptScore W2565257220C153180895 @default.
• W2565257220 hasConceptScore W2565257220C154945302 @default.
• W2565257220 hasConceptScore W2565257220C169805256 @default.
• W2565257220 hasConceptScore W2565257220C198751489 @default.
• W2565257220 hasConceptScore W2565257220C2221639 @default.
• W2565257220 hasConceptScore W2565257220C28855332 @default.
• W2565257220 hasConceptScore W2565257220C41008148 @default.
• W2565257220 hasConceptScore W2565257220C45347329 @default.
• W2565257220 hasConceptScore W2565257220C50644808 @default.
• W2565257220 hasConceptScore W2565257220C81363708 @default.
• W2565257220 hasFunder F4320321001 @default.
• W2565257220 hasIssue "5" @default.
• W2565257220 hasLocation W25652572201 @default.
• W2565257220 hasLocation W25652572202 @default.
• W2565257220 hasOpenAccess W2565257220 @default.
• W2565257220 hasPrimaryLocation W25652572201 @default.
• W2565257220 hasRelatedWork W1529294424 @default.
• W2565257220 hasRelatedWork W1597187555 @default.
• W2565257220 hasRelatedWork W1966340347 @default.
• W2565257220 hasRelatedWork W2037059691 @default.

• next