Matches in SemOpenAlex for { <https://semopenalex.org/work/W3169425223> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W3169425223 abstract "The robustness of deep neural networks against adversarial example attacks has received much attention recently. We focus on certified robustness of smoothed classifiers in this work, and propose to use the worst-case population loss over noisy inputs as a robustness metric. Under this metric, we provide a tractable upper bound serving as a robustness certificate by exploiting the duality. To improve the robustness, we further propose a noisy adversarial learning procedure to minimize the upper bound following the robust optimization framework. The smoothness of the loss function ensures the problem easy to optimize even for non-smooth neural networks. We show how our robustness certificate compares with others and the improvement over previous works. Experiments on a variety of datasets and models verify that in terms of empirical accuracies, our approach exceeds the state-of-the-art certified/heuristic methods in defending adversarial examples." @default.
- W3169425223 created "2021-06-22" @default.
- W3169425223 creator A5018175815 @default.
- W3169425223 creator A5034483183 @default.
- W3169425223 creator A5045198205 @default.
- W3169425223 creator A5046597133 @default.
- W3169425223 creator A5061582301 @default.
- W3169425223 creator A5072326707 @default.
- W3169425223 date "2021-05-04" @default.
- W3169425223 modified "2023-09-26" @default.
- W3169425223 title "Certified Distributional Robustness via Smoothed Classifiers" @default.
- W3169425223 cites W1484551447 @default.
- W3169425223 cites W2342045095 @default.
- W3169425223 cites W2552767274 @default.
- W3169425223 cites W2766462876 @default.
- W3169425223 cites W2767075075 @default.
- W3169425223 cites W2791953061 @default.
- W3169425223 cites W2803850896 @default.
- W3169425223 cites W2807040120 @default.
- W3169425223 cites W2892354372 @default.
- W3169425223 cites W2898963688 @default.
- W3169425223 cites W2913266441 @default.
- W3169425223 cites W2913848079 @default.
- W3169425223 cites W2944407464 @default.
- W3169425223 cites W2950048339 @default.
- W3169425223 cites W2962710014 @default.
- W3169425223 cites W2963207607 @default.
- W3169425223 cites W2963450292 @default.
- W3169425223 cites W2963496101 @default.
- W3169425223 cites W2963626858 @default.
- W3169425223 cites W2963952467 @default.
- W3169425223 cites W2964253222 @default.
- W3169425223 cites W2964814686 @default.
- W3169425223 cites W2970209211 @default.
- W3169425223 cites W2970297683 @default.
- W3169425223 cites W2971109239 @default.
- W3169425223 cites W2989696285 @default.
- W3169425223 cites W2990740676 @default.
- W3169425223 cites W2995053405 @default.
- W3169425223 cites W2996344901 @default.
- W3169425223 cites W3006335454 @default.
- W3169425223 cites W3034936311 @default.
- W3169425223 hasPublicationYear "2021" @default.
- W3169425223 type Work @default.
- W3169425223 sameAs 3169425223 @default.
- W3169425223 citedByCount "0" @default.
- W3169425223 crossrefType "journal-article" @default.
- W3169425223 hasAuthorship W3169425223A5018175815 @default.
- W3169425223 hasAuthorship W3169425223A5034483183 @default.
- W3169425223 hasAuthorship W3169425223A5045198205 @default.
- W3169425223 hasAuthorship W3169425223A5046597133 @default.
- W3169425223 hasAuthorship W3169425223A5061582301 @default.
- W3169425223 hasAuthorship W3169425223A5072326707 @default.
- W3169425223 hasConcept C104317684 @default.
- W3169425223 hasConcept C119857082 @default.
- W3169425223 hasConcept C126255220 @default.
- W3169425223 hasConcept C134306372 @default.
- W3169425223 hasConcept C154945302 @default.
- W3169425223 hasConcept C185592680 @default.
- W3169425223 hasConcept C33923547 @default.
- W3169425223 hasConcept C41008148 @default.
- W3169425223 hasConcept C55493867 @default.
- W3169425223 hasConcept C63479239 @default.
- W3169425223 hasConcept C77553402 @default.
- W3169425223 hasConceptScore W3169425223C104317684 @default.
- W3169425223 hasConceptScore W3169425223C119857082 @default.
- W3169425223 hasConceptScore W3169425223C126255220 @default.
- W3169425223 hasConceptScore W3169425223C134306372 @default.
- W3169425223 hasConceptScore W3169425223C154945302 @default.
- W3169425223 hasConceptScore W3169425223C185592680 @default.
- W3169425223 hasConceptScore W3169425223C33923547 @default.
- W3169425223 hasConceptScore W3169425223C41008148 @default.
- W3169425223 hasConceptScore W3169425223C55493867 @default.
- W3169425223 hasConceptScore W3169425223C63479239 @default.
- W3169425223 hasConceptScore W3169425223C77553402 @default.
- W3169425223 hasOpenAccess W3169425223 @default.
- W3169425223 hasRelatedWork W2395317528 @default.
- W3169425223 hasRelatedWork W2897959453 @default.
- W3169425223 hasRelatedWork W2962756933 @default.
- W3169425223 hasRelatedWork W2964540166 @default.
- W3169425223 hasRelatedWork W2979999556 @default.
- W3169425223 hasRelatedWork W2981817257 @default.
- W3169425223 hasRelatedWork W2982615324 @default.
- W3169425223 hasRelatedWork W2987875759 @default.
- W3169425223 hasRelatedWork W3019321885 @default.
- W3169425223 hasRelatedWork W3028763319 @default.
- W3169425223 hasRelatedWork W3033323705 @default.
- W3169425223 hasRelatedWork W3035168486 @default.
- W3169425223 hasRelatedWork W3091918696 @default.
- W3169425223 hasRelatedWork W3094319733 @default.
- W3169425223 hasRelatedWork W3112131305 @default.
- W3169425223 hasRelatedWork W3128448405 @default.
- W3169425223 hasRelatedWork W3135151133 @default.
- W3169425223 hasRelatedWork W3168586460 @default.
- W3169425223 hasRelatedWork W3175117184 @default.
- W3169425223 hasRelatedWork W3190398575 @default.
- W3169425223 isParatext "false" @default.
- W3169425223 isRetracted "false" @default.
- W3169425223 magId "3169425223" @default.
- W3169425223 workType "article" @default.