FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2970456043> ?p ?o ?g. }

• W2970456043 endingPage "15244" @default.
• W2970456043 startingPage "15234" @default.
• W2970456043 abstract "The use of neural networks in safety-critical computer vision systems calls for their robustness certification against natural geometric transformations (e.g., rotation, scaling). However, current certification methods target mostly norm-based pixel perturbations and cannot certify robustness against geometric transformations. In this work, we propose a new method to compute sound and asymptotically optimal linear relaxations for any composition of transformations. Our method is based on a novel combination of sampling and optimization. We implemented the method in a system called DeepG and demonstrated that it certifies significantly more complex geometric transformations than existing methods on both defended and undefended networks while scaling to large architectures." @default.
• W2970456043 created "2019-09-05" @default.
• W2970456043 creator A5014657381 @default.
• W2970456043 creator A5018338894 @default.
• W2970456043 creator A5045044745 @default.
• W2970456043 creator A5052782390 @default.
• W2970456043 creator A5069901599 @default.
• W2970456043 date "2019-01-01" @default.
• W2970456043 modified "2023-09-29" @default.
• W2970456043 title "Certifying Geometric Robustness of Neural Networks" @default.
• W2970456043 cites W2105712817 @default.
• W2970456043 cites W2105728138 @default.
• W2970456043 cites W2132661148 @default.
• W2970456043 cites W2566919497 @default.
• W2970456043 cites W2594877703 @default.
• W2970456043 cites W2750384547 @default.
• W2970456043 cites W2767471303 @default.
• W2970456043 cites W2768718880 @default.
• W2970456043 cites W2783784437 @default.
• W2970456043 cites W2794609696 @default.
• W2970456043 cites W2800473699 @default.
• W2970456043 cites W2803850896 @default.
• W2970456043 cites W2807040120 @default.
• W2970456043 cites W2890660842 @default.
• W2970456043 cites W2892354372 @default.
• W2970456043 cites W2900153411 @default.
• W2970456043 cites W2945033152 @default.
• W2970456043 cites W2963496101 @default.
• W2970456043 cites W2963735478 @default.
• W2970456043 cites W2963960272 @default.
• W2970456043 cites W2964153729 @default.
• W2970456043 cites W2964253222 @default.
• W2970456043 cites W2966661 @default.
• W2970456043 cites W2970297683 @default.
• W2970456043 cites W3118608800 @default.
• W2970456043 cites W603908379 @default.
• W2970456043 cites W9657784 @default.
• W2970456043 doi "https://doi.org/10.3929/ethz-b-000395340" @default.
• W2970456043 hasPublicationYear "2019" @default.
• W2970456043 type Work @default.
• W2970456043 sameAs 2970456043 @default.
• W2970456043 citedByCount "36" @default.
• W2970456043 countsByYear W29704560432019 @default.
• W2970456043 countsByYear W29704560432020 @default.
• W2970456043 countsByYear W29704560432021 @default.
• W2970456043 countsByYear W29704560432022 @default.
• W2970456043 countsByYear W29704560432023 @default.
• W2970456043 crossrefType "proceedings-article" @default.
• W2970456043 hasAuthorship W2970456043A5014657381 @default.
• W2970456043 hasAuthorship W2970456043A5018338894 @default.
• W2970456043 hasAuthorship W2970456043A5045044745 @default.
• W2970456043 hasAuthorship W2970456043A5052782390 @default.
• W2970456043 hasAuthorship W2970456043A5069901599 @default.
• W2970456043 hasConcept C104317684 @default.
• W2970456043 hasConcept C11413529 @default.
• W2970456043 hasConcept C126255220 @default.
• W2970456043 hasConcept C154945302 @default.
• W2970456043 hasConcept C160633673 @default.
• W2970456043 hasConcept C17744445 @default.
• W2970456043 hasConcept C185592680 @default.
• W2970456043 hasConcept C199539241 @default.
• W2970456043 hasConcept C2524010 @default.
• W2970456043 hasConcept C32990609 @default.
• W2970456043 hasConcept C33923547 @default.
• W2970456043 hasConcept C41008148 @default.
• W2970456043 hasConcept C46304622 @default.
• W2970456043 hasConcept C50644808 @default.
• W2970456043 hasConcept C55493867 @default.
• W2970456043 hasConcept C63479239 @default.
• W2970456043 hasConcept C80444323 @default.
• W2970456043 hasConcept C99844830 @default.
• W2970456043 hasConceptScore W2970456043C104317684 @default.
• W2970456043 hasConceptScore W2970456043C11413529 @default.
• W2970456043 hasConceptScore W2970456043C126255220 @default.
• W2970456043 hasConceptScore W2970456043C154945302 @default.
• W2970456043 hasConceptScore W2970456043C160633673 @default.
• W2970456043 hasConceptScore W2970456043C17744445 @default.
• W2970456043 hasConceptScore W2970456043C185592680 @default.
• W2970456043 hasConceptScore W2970456043C199539241 @default.
• W2970456043 hasConceptScore W2970456043C2524010 @default.
• W2970456043 hasConceptScore W2970456043C32990609 @default.
• W2970456043 hasConceptScore W2970456043C33923547 @default.
• W2970456043 hasConceptScore W2970456043C41008148 @default.
• W2970456043 hasConceptScore W2970456043C46304622 @default.
• W2970456043 hasConceptScore W2970456043C50644808 @default.
• W2970456043 hasConceptScore W2970456043C55493867 @default.
• W2970456043 hasConceptScore W2970456043C63479239 @default.
• W2970456043 hasConceptScore W2970456043C80444323 @default.
• W2970456043 hasConceptScore W2970456043C99844830 @default.
• W2970456043 hasLocation W29704560431 @default.
• W2970456043 hasOpenAccess W2970456043 @default.
• W2970456043 hasPrimaryLocation W29704560431 @default.
• W2970456043 hasRelatedWork W2194775991 @default.
• W2970456043 hasRelatedWork W2594877703 @default.
• W2970456043 hasRelatedWork W2794609696 @default.
• W2970456043 hasRelatedWork W2803850896 @default.
• W2970456043 hasRelatedWork W2807040120 @default.
• W2970456043 hasRelatedWork W2890472662 @default.

• next