FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3040473363> ?p ?o ?g. }

• W3040473363 abstract "Convolution neural networks have achieved remarkable performance in many tasks of computing vision. However, CNN tends to bias to low frequency components. They prioritize capturing low frequency patterns which lead them fail when suffering from application scenario transformation. While adversarial example implies the model is very sensitive to high frequency perturbations. In this paper, we introduce a new regularization method by constraining the frequency spectra of the filter of the model. Different from band-limit training, our method considers the valid frequency range probably entangles in different layers rather than continuous and trains the valid frequency range end-to-end by backpropagation. We demonstrate the effectiveness of our regularization by (1) defensing to adversarial perturbations; (2) reducing the generalization gap in different architecture; (3) improving the generalization ability in transfer learning scenario without fine-tune." @default.
• W3040473363 created "2020-07-10" @default.
• W3040473363 creator A5005136271 @default.
• W3040473363 creator A5038164924 @default.
• W3040473363 date "2020-07-07" @default.
• W3040473363 modified "2023-09-26" @default.
• W3040473363 title "Robust Learning with Frequency Domain Regularization" @default.
• W3040473363 cites W1901129140 @default.
• W3040473363 cites W2099471712 @default.
• W3040473363 cites W2144513243 @default.
• W3040473363 cites W2154579312 @default.
• W3040473363 cites W2167034998 @default.
• W3040473363 cites W2194775991 @default.
• W3040473363 cites W2295107390 @default.
• W3040473363 cites W2382313035 @default.
• W3040473363 cites W2612573399 @default.
• W3040473363 cites W2768282280 @default.
• W3040473363 cites W2774616426 @default.
• W3040473363 cites W2787442904 @default.
• W3040473363 cites W2796347433 @default.
• W3040473363 cites W2804047946 @default.
• W3040473363 cites W2807007689 @default.
• W3040473363 cites W2890166761 @default.
• W3040473363 cites W2890304135 @default.
• W3040473363 cites W2900578291 @default.
• W3040473363 cites W2946070613 @default.
• W3040473363 cites W2949117887 @default.
• W3040473363 cites W2949582403 @default.
• W3040473363 cites W2953106684 @default.
• W3040473363 cites W2963000224 @default.
• W3040473363 cites W2963190567 @default.
• W3040473363 cites W2963340555 @default.
• W3040473363 cites W2963367891 @default.
• W3040473363 cites W2963417375 @default.
• W3040473363 cites W2963426391 @default.
• W3040473363 cites W2963494934 @default.
• W3040473363 cites W2963686907 @default.
• W3040473363 cites W2963739978 @default.
• W3040473363 cites W2963775761 @default.
• W3040473363 cites W2963836885 @default.
• W3040473363 cites W2963938169 @default.
• W3040473363 cites W2964294232 @default.
• W3040473363 cites W2964582580 @default.
• W3040473363 cites W2971028215 @default.
• W3040473363 cites W2979544394 @default.
• W3040473363 cites W2994931756 @default.
• W3040473363 cites W3012704771 @default.
• W3040473363 cites W3106250896 @default.
• W3040473363 cites W3118608800 @default.
• W3040473363 cites W3137695714 @default.
• W3040473363 cites W3016625911 @default.
• W3040473363 doi "https://doi.org/10.48550/arxiv.2007.03244" @default.
• W3040473363 hasPublicationYear "2020" @default.
• W3040473363 type Work @default.
• W3040473363 sameAs 3040473363 @default.
• W3040473363 citedByCount "0" @default.
• W3040473363 crossrefType "posted-content" @default.
• W3040473363 hasAuthorship W3040473363A5005136271 @default.
• W3040473363 hasAuthorship W3040473363A5038164924 @default.
• W3040473363 hasBestOaLocation W30404733631 @default.
• W3040473363 hasConcept C104892082 @default.
• W3040473363 hasConcept C11413529 @default.
• W3040473363 hasConcept C134306372 @default.
• W3040473363 hasConcept C154945302 @default.
• W3040473363 hasConcept C155032097 @default.
• W3040473363 hasConcept C159985019 @default.
• W3040473363 hasConcept C177148314 @default.
• W3040473363 hasConcept C19118579 @default.
• W3040473363 hasConcept C192562407 @default.
• W3040473363 hasConcept C204323151 @default.
• W3040473363 hasConcept C2776135515 @default.
• W3040473363 hasConcept C2776257435 @default.
• W3040473363 hasConcept C2778116611 @default.
• W3040473363 hasConcept C31972630 @default.
• W3040473363 hasConcept C33923547 @default.
• W3040473363 hasConcept C37736160 @default.
• W3040473363 hasConcept C41008148 @default.
• W3040473363 hasConcept C45347329 @default.
• W3040473363 hasConcept C50644808 @default.
• W3040473363 hasConcept C76155785 @default.
• W3040473363 hasConceptScore W3040473363C104892082 @default.
• W3040473363 hasConceptScore W3040473363C11413529 @default.
• W3040473363 hasConceptScore W3040473363C134306372 @default.
• W3040473363 hasConceptScore W3040473363C154945302 @default.
• W3040473363 hasConceptScore W3040473363C155032097 @default.
• W3040473363 hasConceptScore W3040473363C159985019 @default.
• W3040473363 hasConceptScore W3040473363C177148314 @default.
• W3040473363 hasConceptScore W3040473363C19118579 @default.
• W3040473363 hasConceptScore W3040473363C192562407 @default.
• W3040473363 hasConceptScore W3040473363C204323151 @default.
• W3040473363 hasConceptScore W3040473363C2776135515 @default.
• W3040473363 hasConceptScore W3040473363C2776257435 @default.
• W3040473363 hasConceptScore W3040473363C2778116611 @default.
• W3040473363 hasConceptScore W3040473363C31972630 @default.
• W3040473363 hasConceptScore W3040473363C33923547 @default.
• W3040473363 hasConceptScore W3040473363C37736160 @default.
• W3040473363 hasConceptScore W3040473363C41008148 @default.
• W3040473363 hasConceptScore W3040473363C45347329 @default.
• W3040473363 hasConceptScore W3040473363C50644808 @default.
• W3040473363 hasConceptScore W3040473363C76155785 @default.

• next