FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3090717570> ?p ?o ?g. }

• W3090717570 abstract "Deep Neural Networks (DNNs) classifiers performance degrades under adversarial attacks, such attacks are indistinguishably perturbed relative to the original data. Providing robustness to adversarial attacks is an important challenge in DNN training, which has led to extensive research. In this paper, we harden DNN classifiers under the adversarial attacks by regularizing their deep internal representation space with Multi-class Triplet regularization method. This method enables DNN classifier to learn a feature representation that detects similarities between adversarial and clean images and brings similar images close to their original class and pushes dissimilar images away from their false classes. This training process with our Multi-class Triplet regularization method in combination with Gaussian noise injection proves to be more robust in detecting adversarial attacks exceeding that of adversarial training on strong iterative attacks." @default.
• W3090717570 created "2020-10-08" @default.
• W3090717570 creator A5036840299 @default.
• W3090717570 creator A5053070964 @default.
• W3090717570 creator A5063951151 @default.
• W3090717570 creator A5082599078 @default.
• W3090717570 date "2020-01-01" @default.
• W3090717570 modified "2023-09-23" @default.
• W3090717570 title "Multi-Class Triplet Loss With Gaussian Noise for Adversarial Robustness" @default.
• W3090717570 cites W1932198206 @default.
• W3090717570 cites W2112796928 @default.
• W3090717570 cites W2180612164 @default.
• W3090717570 cites W2342045095 @default.
• W3090717570 cites W2408141691 @default.
• W3090717570 cites W2520774990 @default.
• W3090717570 cites W2555897561 @default.
• W3090717570 cites W2591712613 @default.
• W3090717570 cites W2607219512 @default.
• W3090717570 cites W2618043096 @default.
• W3090717570 cites W2787947370 @default.
• W3090717570 cites W2789828921 @default.
• W3090717570 cites W2791953061 @default.
• W3090717570 cites W2908392948 @default.
• W3090717570 cites W2911634294 @default.
• W3090717570 cites W2962972208 @default.
• W3090717570 cites W2963100962 @default.
• W3090717570 cites W2963166243 @default.
• W3090717570 cites W2963207607 @default.
• W3090717570 cites W2963446712 @default.
• W3090717570 cites W2963485691 @default.
• W3090717570 cites W2963542245 @default.
• W3090717570 cites W2963564844 @default.
• W3090717570 cites W2963626858 @default.
• W3090717570 cites W2963775347 @default.
• W3090717570 cites W2963857521 @default.
• W3090717570 cites W2964153729 @default.
• W3090717570 cites W2964197269 @default.
• W3090717570 cites W2964253222 @default.
• W3090717570 cites W2964267298 @default.
• W3090717570 cites W2970088379 @default.
• W3090717570 cites W3015625436 @default.
• W3090717570 cites W3099206234 @default.
• W3090717570 cites W3118608800 @default.
• W3090717570 doi "https://doi.org/10.1109/access.2020.3024244" @default.
• W3090717570 hasPublicationYear "2020" @default.
• W3090717570 type Work @default.
• W3090717570 sameAs 3090717570 @default.
• W3090717570 citedByCount "0" @default.
• W3090717570 crossrefType "journal-article" @default.
• W3090717570 hasAuthorship W3090717570A5036840299 @default.
• W3090717570 hasAuthorship W3090717570A5053070964 @default.
• W3090717570 hasAuthorship W3090717570A5063951151 @default.
• W3090717570 hasAuthorship W3090717570A5082599078 @default.
• W3090717570 hasBestOaLocation W30907175701 @default.
• W3090717570 hasConcept C104317684 @default.
• W3090717570 hasConcept C119857082 @default.
• W3090717570 hasConcept C121332964 @default.
• W3090717570 hasConcept C153180895 @default.
• W3090717570 hasConcept C154945302 @default.
• W3090717570 hasConcept C163716315 @default.
• W3090717570 hasConcept C185592680 @default.
• W3090717570 hasConcept C2776135515 @default.
• W3090717570 hasConcept C2984842247 @default.
• W3090717570 hasConcept C37736160 @default.
• W3090717570 hasConcept C41008148 @default.
• W3090717570 hasConcept C4199805 @default.
• W3090717570 hasConcept C50644808 @default.
• W3090717570 hasConcept C55493867 @default.
• W3090717570 hasConcept C62520636 @default.
• W3090717570 hasConcept C63479239 @default.
• W3090717570 hasConcept C83665646 @default.
• W3090717570 hasConcept C95623464 @default.
• W3090717570 hasConceptScore W3090717570C104317684 @default.
• W3090717570 hasConceptScore W3090717570C119857082 @default.
• W3090717570 hasConceptScore W3090717570C121332964 @default.
• W3090717570 hasConceptScore W3090717570C153180895 @default.
• W3090717570 hasConceptScore W3090717570C154945302 @default.
• W3090717570 hasConceptScore W3090717570C163716315 @default.
• W3090717570 hasConceptScore W3090717570C185592680 @default.
• W3090717570 hasConceptScore W3090717570C2776135515 @default.
• W3090717570 hasConceptScore W3090717570C2984842247 @default.
• W3090717570 hasConceptScore W3090717570C37736160 @default.
• W3090717570 hasConceptScore W3090717570C41008148 @default.
• W3090717570 hasConceptScore W3090717570C4199805 @default.
• W3090717570 hasConceptScore W3090717570C50644808 @default.
• W3090717570 hasConceptScore W3090717570C55493867 @default.
• W3090717570 hasConceptScore W3090717570C62520636 @default.
• W3090717570 hasConceptScore W3090717570C63479239 @default.
• W3090717570 hasConceptScore W3090717570C83665646 @default.
• W3090717570 hasConceptScore W3090717570C95623464 @default.
• W3090717570 hasLocation W30907175701 @default.
• W3090717570 hasOpenAccess W3090717570 @default.
• W3090717570 hasPrimaryLocation W30907175701 @default.
• W3090717570 hasRelatedWork W11300528 @default.
• W3090717570 hasRelatedWork W11389402 @default.
• W3090717570 hasRelatedWork W13426584 @default.
• W3090717570 hasRelatedWork W13435860 @default.
• W3090717570 hasRelatedWork W14481880 @default.
• W3090717570 hasRelatedWork W14536956 @default.
• W3090717570 hasRelatedWork W5477720 @default.

• next