Matches in SemOpenAlex for { <https://semopenalex.org/work/W3101700548> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W3101700548 endingPage "8601" @default.
- W3101700548 startingPage "8588" @default.
- W3101700548 abstract "Current methods for training robust networks lead to a drop in test accuracy, which has led prior works to posit that a robustness-accuracy tradeoff may be inevitable in deep learning. We take a closer look at this phenomenon and first show that real image datasets are actually separated. With this property in mind, we then prove that robustness and accuracy should both be achievable for benchmark datasets through locally Lipschitz functions, and hence, there should be no inherent tradeoff between robustness and accuracy. Through extensive experiments with robustness methods, we argue that the gap between theory and practice arises from two limitations of current methods: either they fail to impose local Lipschitzness or they are insufficiently generalized. We explore combining dropout with robust training methods and obtain better generalization. We conclude that achieving robustness and accuracy in practice may require using methods that impose local Lipschitzness and augmenting them with deep learning generalization techniques. Code available at this https URL" @default.
- W3101700548 created "2020-11-23" @default.
- W3101700548 creator A5010790447 @default.
- W3101700548 creator A5018348928 @default.
- W3101700548 creator A5021842545 @default.
- W3101700548 creator A5033186377 @default.
- W3101700548 creator A5066091408 @default.
- W3101700548 date "2020-01-01" @default.
- W3101700548 modified "2023-09-24" @default.
- W3101700548 title "A Closer Look at Accuracy vs. Robustness" @default.
- W3101700548 hasPublicationYear "2020" @default.
- W3101700548 type Work @default.
- W3101700548 sameAs 3101700548 @default.
- W3101700548 citedByCount "22" @default.
- W3101700548 countsByYear W31017005482020 @default.
- W3101700548 countsByYear W31017005482021 @default.
- W3101700548 countsByYear W31017005482022 @default.
- W3101700548 countsByYear W31017005482023 @default.
- W3101700548 crossrefType "proceedings-article" @default.
- W3101700548 hasAuthorship W3101700548A5010790447 @default.
- W3101700548 hasAuthorship W3101700548A5018348928 @default.
- W3101700548 hasAuthorship W3101700548A5021842545 @default.
- W3101700548 hasAuthorship W3101700548A5033186377 @default.
- W3101700548 hasAuthorship W3101700548A5066091408 @default.
- W3101700548 hasConcept C104317684 @default.
- W3101700548 hasConcept C11413529 @default.
- W3101700548 hasConcept C119857082 @default.
- W3101700548 hasConcept C154945302 @default.
- W3101700548 hasConcept C185592680 @default.
- W3101700548 hasConcept C41008148 @default.
- W3101700548 hasConcept C55493867 @default.
- W3101700548 hasConcept C63479239 @default.
- W3101700548 hasConceptScore W3101700548C104317684 @default.
- W3101700548 hasConceptScore W3101700548C11413529 @default.
- W3101700548 hasConceptScore W3101700548C119857082 @default.
- W3101700548 hasConceptScore W3101700548C154945302 @default.
- W3101700548 hasConceptScore W3101700548C185592680 @default.
- W3101700548 hasConceptScore W3101700548C41008148 @default.
- W3101700548 hasConceptScore W3101700548C55493867 @default.
- W3101700548 hasConceptScore W3101700548C63479239 @default.
- W3101700548 hasLocation W31017005481 @default.
- W3101700548 hasOpenAccess W3101700548 @default.
- W3101700548 hasPrimaryLocation W31017005481 @default.
- W3101700548 hasRelatedWork W2194775991 @default.
- W3101700548 hasRelatedWork W2911634294 @default.
- W3101700548 hasRelatedWork W2962729158 @default.
- W3101700548 hasRelatedWork W2963143631 @default.
- W3101700548 hasRelatedWork W2963207607 @default.
- W3101700548 hasRelatedWork W2963496101 @default.
- W3101700548 hasRelatedWork W2963857521 @default.
- W3101700548 hasRelatedWork W2964116600 @default.
- W3101700548 hasRelatedWork W2964153729 @default.
- W3101700548 hasRelatedWork W2964253222 @default.
- W3101700548 hasRelatedWork W2970115835 @default.
- W3101700548 hasRelatedWork W2996344901 @default.
- W3101700548 hasRelatedWork W3034758058 @default.
- W3101700548 hasRelatedWork W3034994123 @default.
- W3101700548 hasRelatedWork W3035032188 @default.
- W3101700548 hasRelatedWork W3042135962 @default.
- W3101700548 hasRelatedWork W3092171228 @default.
- W3101700548 hasRelatedWork W3103385169 @default.
- W3101700548 hasRelatedWork W3118608800 @default.
- W3101700548 hasRelatedWork W9657784 @default.
- W3101700548 hasVolume "33" @default.
- W3101700548 isParatext "false" @default.
- W3101700548 isRetracted "false" @default.
- W3101700548 magId "3101700548" @default.
- W3101700548 workType "article" @default.