FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3131617510> ?p ?o ?g. }

• W3131617510 abstract "Recent improvements in generative adversarial visual synthesis incorporate real and fake image transformation in a self-supervised setting, leading to increased stability and perceptual fidelity. However, these approaches typically involve image augmentations via additional regularizers in the GAN objective and thus spend valuable network capacity towards approximating transformation equivariance instead of their desired task. In this work, we explicitly incorporate inductive symmetry priors into the network architectures via group-equivariant convolutional networks. Group-convolutions have higher expressive power with fewer samples and lead to better gradient feedback between generator and discriminator. Further, group-representations are especially relevant for modalities such as medical imaging, where image labels are typically invariant under rotation/reflection. We show that group-equivariance integrates seamlessly with recent techniques for GAN training across regularizers, architectures, and loss functions. We demonstrate the utility of our methods for conditional synthesis by improving generation in the limited data regime across globally-symmetric imaging datasets and even finding benefits for natural images with preferred orientation." @default.
• W3131617510 created "2021-03-01" @default.
• W3131617510 creator A5036328196 @default.
• W3131617510 creator A5036730225 @default.
• W3131617510 creator A5062044760 @default.
• W3131617510 date "2021-05-03" @default.
• W3131617510 modified "2023-09-28" @default.
• W3131617510 title "Group Equivariant Generative Adversarial Networks" @default.
• W3131617510 cites W12634471 @default.
• W3131617510 cites W1686810756 @default.
• W3131617510 cites W1994197834 @default.
• W3131617510 cites W1994906459 @default.
• W3131617510 cites W2027037828 @default.
• W3131617510 cites W2072072671 @default.
• W3131617510 cites W2099471712 @default.
• W3131617510 cites W2167383966 @default.
• W3131617510 cites W2173520492 @default.
• W3131617510 cites W2183341477 @default.
• W3131617510 cites W2279221249 @default.
• W3131617510 cites W2567627528 @default.
• W3131617510 cites W2768793812 @default.
• W3131617510 cites W2804078698 @default.
• W3131617510 cites W2806140208 @default.
• W3131617510 cites W2806857275 @default.
• W3131617510 cites W2810518847 @default.
• W3131617510 cites W2884013187 @default.
• W3131617510 cites W2890054847 @default.
• W3131617510 cites W2893749619 @default.
• W3131617510 cites W2913314773 @default.
• W3131617510 cites W2922190387 @default.
• W3131617510 cites W2931742923 @default.
• W3131617510 cites W2949117887 @default.
• W3131617510 cites W2951770173 @default.
• W3131617510 cites W2953030256 @default.
• W3131617510 cites W2953327099 @default.
• W3131617510 cites W2962754210 @default.
• W3131617510 cites W2962760235 @default.
• W3131617510 cites W2962879692 @default.
• W3131617510 cites W2962919088 @default.
• W3131617510 cites W2963073614 @default.
• W3131617510 cites W2963143796 @default.
• W3131617510 cites W2963439970 @default.
• W3131617510 cites W2963476931 @default.
• W3131617510 cites W2963675401 @default.
• W3131617510 cites W2963703618 @default.
• W3131617510 cites W2963836885 @default.
• W3131617510 cites W2963921132 @default.
• W3131617510 cites W2963977471 @default.
• W3131617510 cites W2963981733 @default.
• W3131617510 cites W2964121744 @default.
• W3131617510 cites W2964259506 @default.
• W3131617510 cites W2966661 @default.
• W3131617510 cites W2970754912 @default.
• W3131617510 cites W2970756352 @default.
• W3131617510 cites W2971128425 @default.
• W3131617510 cites W2979776030 @default.
• W3131617510 cites W2982511758 @default.
• W3131617510 cites W2984648456 @default.
• W3131617510 cites W2989736850 @default.
• W3131617510 cites W2990706204 @default.
• W3131617510 cites W2995289379 @default.
• W3131617510 cites W2995910740 @default.
• W3131617510 cites W2996690341 @default.
• W3131617510 cites W3004988234 @default.
• W3131617510 cites W3005601278 @default.
• W3131617510 cites W3007166982 @default.
• W3131617510 cites W3008814843 @default.
• W3131617510 cites W3015694949 @default.
• W3131617510 cites W3035574324 @default.
• W3131617510 cites W3080350040 @default.
• W3131617510 cites W3118608800 @default.
• W3131617510 hasPublicationYear "2021" @default.
• W3131617510 type Work @default.
• W3131617510 sameAs 3131617510 @default.
• W3131617510 citedByCount "4" @default.
• W3131617510 countsByYear W31316175102021 @default.
• W3131617510 crossrefType "proceedings-article" @default.
• W3131617510 hasAuthorship W3131617510A5036328196 @default.
• W3131617510 hasAuthorship W3131617510A5036730225 @default.
• W3131617510 hasAuthorship W3131617510A5062044760 @default.
• W3131617510 hasConcept C121332964 @default.
• W3131617510 hasConcept C153180895 @default.
• W3131617510 hasConcept C154945302 @default.
• W3131617510 hasConcept C163258240 @default.
• W3131617510 hasConcept C171036898 @default.
• W3131617510 hasConcept C190470478 @default.
• W3131617510 hasConcept C202444582 @default.
• W3131617510 hasConcept C2779803651 @default.
• W3131617510 hasConcept C2780992000 @default.
• W3131617510 hasConcept C33923547 @default.
• W3131617510 hasConcept C37914503 @default.
• W3131617510 hasConcept C41008148 @default.
• W3131617510 hasConcept C62520636 @default.
• W3131617510 hasConcept C76155785 @default.
• W3131617510 hasConcept C94915269 @default.
• W3131617510 hasConceptScore W3131617510C121332964 @default.
• W3131617510 hasConceptScore W3131617510C153180895 @default.
• W3131617510 hasConceptScore W3131617510C154945302 @default.
• W3131617510 hasConceptScore W3131617510C163258240 @default.
• W3131617510 hasConceptScore W3131617510C171036898 @default.

• next