FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4361805434> ?p ?o ?g. }

• W4361805434 endingPage "2233" @default.
• W4361805434 startingPage "2218" @default.
• W4361805434 abstract "We propose a new approach for privacy-preserving and verifiable convolutional neural network (CNN) testing in a distrustful multi-stakeholder environment. The approach is aimed to enable that a CNN model <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>developer</i> convinces a <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>user</i> of the truthful CNN performance over non-public data from <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>multiple testers</i> , while respecting model and data privacy. To balance the security and efficiency issues, we appropriately integrate three tools with the CNN testing, including collaborative inference, homomorphic encryption (HE) and zero-knowledge succinct non-interactive argument of knowledge (zk-SNARK). We start with strategically partitioning a CNN model into a private part kept locally by the model developer, and a public part outsourced to an outside server. Then, the private part runs over the HE-protected test data sent by a tester, and transmits its outputs to the public part for accomplishing subsequent computations of the CNN testing. Second, the correctness of the above CNN testing is enforced by generating zk-SNARK based proofs, with an emphasis on optimizing proving overhead for two-dimensional (2-D) convolution operations, since the operations dominate the performance bottleneck during generating proofs. We specifically present a new quadratic matrix program (QMP)-based arithmetic circuit with <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>a single multiplication gate</i> for expressing 2-D convolution operations between multiple filters and inputs in a batch manner. Third, we aggregate multiple proofs with respect to a same CNN model but different testers’ test data ( <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>i.e</i> ., different statements) into one proof, and ensure that the validity of the aggregated proof implies the validity of the original multiple proofs. Lastly, our experimental results demonstrate that our QMP-based zk-SNARK performs nearly 13.9× faster than the existing quadratic arithmetic program (QAP)-based zk-SNARK in proving time, and 17.6× faster in Setup time, for high-dimension matrix multiplication. Besides, the limitation on handling a bounded number of multiplications of QAP-based zk-SNARK is relieved." @default.
• W4361805434 created "2023-04-05" @default.
• W4361805434 creator A5000372510 @default.
• W4361805434 creator A5008349813 @default.
• W4361805434 creator A5025910280 @default.
• W4361805434 creator A5059067637 @default.
• W4361805434 creator A5082041657 @default.
• W4361805434 creator A5086862422 @default.
• W4361805434 date "2023-01-01" @default.
• W4361805434 modified "2023-09-30" @default.
• W4361805434 title "pvCNN: Privacy-Preserving and Verifiable Convolutional Neural Network Testing" @default.
• W4361805434 cites W2071520502 @default.
• W4361805434 cites W2147800946 @default.
• W4361805434 cites W2326245332 @default.
• W4361805434 cites W2405356014 @default.
• W4361805434 cites W2496543269 @default.
• W4361805434 cites W2701059868 @default.
• W4361805434 cites W2781091734 @default.
• W4361805434 cites W2791459719 @default.
• W4361805434 cites W2811014381 @default.
• W4361805434 cites W2882986534 @default.
• W4361805434 cites W2886845474 @default.
• W4361805434 cites W2948833786 @default.
• W4361805434 cites W2963208512 @default.
• W4361805434 cites W2963540401 @default.
• W4361805434 cites W2964279767 @default.
• W4361805434 cites W2989230924 @default.
• W4361805434 cites W2989885118 @default.
• W4361805434 cites W3003532255 @default.
• W4361805434 cites W3016063723 @default.
• W4361805434 cites W3023051856 @default.
• W4361805434 cites W3033311948 @default.
• W4361805434 cites W3096076085 @default.
• W4361805434 cites W3096609273 @default.
• W4361805434 cites W3097272009 @default.
• W4361805434 cites W3099444373 @default.
• W4361805434 cites W3111925745 @default.
• W4361805434 cites W3135955057 @default.
• W4361805434 cites W3157453869 @default.
• W4361805434 cites W3161274233 @default.
• W4361805434 cites W3164110928 @default.
• W4361805434 cites W3194922745 @default.
• W4361805434 cites W3213013974 @default.
• W4361805434 cites W3214716770 @default.
• W4361805434 cites W4324007070 @default.
• W4361805434 doi "https://doi.org/10.1109/tifs.2023.3262932" @default.
• W4361805434 hasPublicationYear "2023" @default.
• W4361805434 type Work @default.
• W4361805434 citedByCount "1" @default.
• W4361805434 countsByYear W43618054342023 @default.
• W4361805434 crossrefType "journal-article" @default.
• W4361805434 hasAuthorship W4361805434A5000372510 @default.
• W4361805434 hasAuthorship W4361805434A5008349813 @default.
• W4361805434 hasAuthorship W4361805434A5025910280 @default.
• W4361805434 hasAuthorship W4361805434A5059067637 @default.
• W4361805434 hasAuthorship W4361805434A5082041657 @default.
• W4361805434 hasAuthorship W4361805434A5086862422 @default.
• W4361805434 hasBestOaLocation W43618054342 @default.
• W4361805434 hasConcept C108710211 @default.
• W4361805434 hasConcept C11413529 @default.
• W4361805434 hasConcept C148730421 @default.
• W4361805434 hasConcept C154945302 @default.
• W4361805434 hasConcept C158338273 @default.
• W4361805434 hasConcept C177264268 @default.
• W4361805434 hasConcept C199360897 @default.
• W4361805434 hasConcept C2524010 @default.
• W4361805434 hasConcept C33923547 @default.
• W4361805434 hasConcept C38652104 @default.
• W4361805434 hasConcept C41008148 @default.
• W4361805434 hasConcept C55439883 @default.
• W4361805434 hasConcept C80444323 @default.
• W4361805434 hasConcept C81363708 @default.
• W4361805434 hasConcept C85847156 @default.
• W4361805434 hasConceptScore W4361805434C108710211 @default.
• W4361805434 hasConceptScore W4361805434C11413529 @default.
• W4361805434 hasConceptScore W4361805434C148730421 @default.
• W4361805434 hasConceptScore W4361805434C154945302 @default.
• W4361805434 hasConceptScore W4361805434C158338273 @default.
• W4361805434 hasConceptScore W4361805434C177264268 @default.
• W4361805434 hasConceptScore W4361805434C199360897 @default.
• W4361805434 hasConceptScore W4361805434C2524010 @default.
• W4361805434 hasConceptScore W4361805434C33923547 @default.
• W4361805434 hasConceptScore W4361805434C38652104 @default.
• W4361805434 hasConceptScore W4361805434C41008148 @default.
• W4361805434 hasConceptScore W4361805434C55439883 @default.
• W4361805434 hasConceptScore W4361805434C80444323 @default.
• W4361805434 hasConceptScore W4361805434C81363708 @default.
• W4361805434 hasConceptScore W4361805434C85847156 @default.
• W4361805434 hasFunder F4320321001 @default.
• W4361805434 hasFunder F4320335777 @default.
• W4361805434 hasLocation W43618054341 @default.
• W4361805434 hasLocation W43618054342 @default.
• W4361805434 hasOpenAccess W4361805434 @default.
• W4361805434 hasPrimaryLocation W43618054341 @default.
• W4361805434 hasRelatedWork W1586131529 @default.
• W4361805434 hasRelatedWork W1676299812 @default.
• W4361805434 hasRelatedWork W1802540854 @default.
• W4361805434 hasRelatedWork W2232878900 @default.

• next