Matches in SemOpenAlex for { <https://semopenalex.org/work/W2914520121> ?p ?o ?g. }
- W2914520121 abstract "The principle of equivariance to symmetry transformations enables a theoretically grounded approach to neural network architecture design. Equivariant networks have shown excellent performance and data efficiency on vision and medical imaging problems that exhibit symmetries. Here we show how this principle can be extended beyond global symmetries to local gauge transformations. This enables the development of a very general class of convolutional neural networks on manifolds that depend only on the intrinsic geometry, and which includes many popular methods from equivariant and geometric deep learning. We implement gauge equivariant CNNs for signals defined on the surface of the icosahedron, which provides a reasonable approximation of the sphere. By choosing to work with this very regular manifold, we are able to implement the gauge equivariant convolution using a single conv2d call, making it a highly scalable and practical alternative to Spherical CNNs. Using this method, we demonstrate substantial improvements over previous methods on the task of segmenting omnidirectional images and global climate patterns." @default.
- W2914520121 created "2019-02-21" @default.
- W2914520121 creator A5059215520 @default.
- W2914520121 creator A5083922034 @default.
- W2914520121 creator A5087368991 @default.
- W2914520121 creator A5088933660 @default.
- W2914520121 date "2019-02-11" @default.
- W2914520121 modified "2023-09-27" @default.
- W2914520121 title "Gauge Equivariant Convolutional Networks and the Icosahedral CNN" @default.
- W2914520121 cites W1509036934 @default.
- W2914520121 cites W1901129140 @default.
- W2914520121 cites W1932922755 @default.
- W2914520121 cites W1951806617 @default.
- W2914520121 cites W21223778 @default.
- W2914520121 cites W2194775991 @default.
- W2914520121 cites W2229156033 @default.
- W2914520121 cites W2279221249 @default.
- W2914520121 cites W2558748708 @default.
- W2914520121 cites W2569680626 @default.
- W2914520121 cites W2576915720 @default.
- W2914520121 cites W2586114507 @default.
- W2914520121 cites W2594345251 @default.
- W2914520121 cites W2751706698 @default.
- W2914520121 cites W2769152880 @default.
- W2914520121 cites W2785397276 @default.
- W2914520121 cites W2787219183 @default.
- W2914520121 cites W2788775653 @default.
- W2914520121 cites W2795221267 @default.
- W2914520121 cites W2796422723 @default.
- W2914520121 cites W2797461639 @default.
- W2914520121 cites W2803679278 @default.
- W2914520121 cites W2806857275 @default.
- W2914520121 cites W2810026216 @default.
- W2914520121 cites W2898187790 @default.
- W2914520121 cites W2900364424 @default.
- W2914520121 cites W2906340703 @default.
- W2914520121 cites W2906915746 @default.
- W2914520121 cites W2906991313 @default.
- W2914520121 cites W2909320927 @default.
- W2914520121 cites W2949117887 @default.
- W2914520121 cites W2949976921 @default.
- W2914520121 cites W2963021451 @default.
- W2914520121 cites W2963343946 @default.
- W2914520121 cites W2963425704 @default.
- W2914520121 cites W2963527880 @default.
- W2914520121 cites W2963737468 @default.
- W2914520121 cites W2963887288 @default.
- W2914520121 cites W2963991385 @default.
- W2914520121 cites W2964015378 @default.
- W2914520121 cites W2964213081 @default.
- W2914520121 cites W2964311892 @default.
- W2914520121 cites W2979240068 @default.
- W2914520121 cites W3138516063 @default.
- W2914520121 cites W72050942 @default.
- W2914520121 hasPublicationYear "2019" @default.
- W2914520121 type Work @default.
- W2914520121 sameAs 2914520121 @default.
- W2914520121 citedByCount "60" @default.
- W2914520121 countsByYear W29145201212018 @default.
- W2914520121 countsByYear W29145201212019 @default.
- W2914520121 countsByYear W29145201212020 @default.
- W2914520121 countsByYear W29145201212021 @default.
- W2914520121 countsByYear W29145201212022 @default.
- W2914520121 countsByYear W29145201212023 @default.
- W2914520121 crossrefType "posted-content" @default.
- W2914520121 hasAuthorship W2914520121A5059215520 @default.
- W2914520121 hasAuthorship W2914520121A5083922034 @default.
- W2914520121 hasAuthorship W2914520121A5087368991 @default.
- W2914520121 hasAuthorship W2914520121A5088933660 @default.
- W2914520121 hasConcept C108583219 @default.
- W2914520121 hasConcept C114614502 @default.
- W2914520121 hasConcept C127413603 @default.
- W2914520121 hasConcept C154945302 @default.
- W2914520121 hasConcept C171036898 @default.
- W2914520121 hasConcept C184720557 @default.
- W2914520121 hasConcept C202444582 @default.
- W2914520121 hasConcept C2524010 @default.
- W2914520121 hasConcept C2779886137 @default.
- W2914520121 hasConcept C33923547 @default.
- W2914520121 hasConcept C41008148 @default.
- W2914520121 hasConcept C529865628 @default.
- W2914520121 hasConcept C78519656 @default.
- W2914520121 hasConcept C81363708 @default.
- W2914520121 hasConcept C96469262 @default.
- W2914520121 hasConceptScore W2914520121C108583219 @default.
- W2914520121 hasConceptScore W2914520121C114614502 @default.
- W2914520121 hasConceptScore W2914520121C127413603 @default.
- W2914520121 hasConceptScore W2914520121C154945302 @default.
- W2914520121 hasConceptScore W2914520121C171036898 @default.
- W2914520121 hasConceptScore W2914520121C184720557 @default.
- W2914520121 hasConceptScore W2914520121C202444582 @default.
- W2914520121 hasConceptScore W2914520121C2524010 @default.
- W2914520121 hasConceptScore W2914520121C2779886137 @default.
- W2914520121 hasConceptScore W2914520121C33923547 @default.
- W2914520121 hasConceptScore W2914520121C41008148 @default.
- W2914520121 hasConceptScore W2914520121C529865628 @default.
- W2914520121 hasConceptScore W2914520121C78519656 @default.
- W2914520121 hasConceptScore W2914520121C81363708 @default.
- W2914520121 hasConceptScore W2914520121C96469262 @default.
- W2914520121 hasLocation W29145201211 @default.