Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313002291> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W4313002291 endingPage "526" @default.
- W4313002291 startingPage "510" @default.
- W4313002291 abstract "We present a method for finding paths in a deep generative model’s latent space that can maximally vary one set of image features while holding others constant. Crucially, unlike past traversal approaches, ours can manipulate arbitrary multidimensional features of an image such as facial identity and pixels within a specified region. Our method is principled and conceptually simple: optimal traversal directions are chosen by maximizing differential changes to one feature set such that changes to another set are negligible. We show that this problem is nearly equivalent to one of Rayleigh quotient maximization, and provide a closed-form solution to it based on solving a generalized eigenvalue equation. We use repeated computations of the corresponding optimal directions, which we call Rayleigh EigenDirections (REDs), to generate appropriately curved paths in latent space. We empirically evaluate our method using StyleGAN2 and BigGAN on the following image domains: faces, living rooms and ImageNet. We show that our method is capable of controlling various multidimensional features: face identity, geometric and semantic attributes, spatial frequency bands, pixels within a region, and the appearance and position of an object. Our work suggests that a wealth of opportunities lies in the local analysis of the geometry and semantics of latent spaces." @default.
- W4313002291 created "2023-01-05" @default.
- W4313002291 creator A5015956846 @default.
- W4313002291 creator A5026825380 @default.
- W4313002291 creator A5081710525 @default.
- W4313002291 creator A5085208414 @default.
- W4313002291 date "2022-01-01" @default.
- W4313002291 modified "2023-10-14" @default.
- W4313002291 title "Rayleigh EigenDirections (REDs): Nonlinear GAN Latent Space Traversals for Multidimensional Features" @default.
- W4313002291 cites W2587706859 @default.
- W4313002291 cites W2737047298 @default.
- W4313002291 cites W2799209711 @default.
- W4313002291 cites W2939645724 @default.
- W4313002291 cites W2962770929 @default.
- W4313002291 cites W2962974533 @default.
- W4313002291 cites W2963100452 @default.
- W4313002291 cites W2963626105 @default.
- W4313002291 cites W2963767194 @default.
- W4313002291 cites W2963800363 @default.
- W4313002291 cites W2963870144 @default.
- W4313002291 cites W2964118024 @default.
- W4313002291 cites W2964231450 @default.
- W4313002291 cites W2964339532 @default.
- W4313002291 cites W2969985801 @default.
- W4313002291 cites W3003162010 @default.
- W4313002291 cites W3034431451 @default.
- W4313002291 cites W3035574324 @default.
- W4313002291 cites W3108663077 @default.
- W4313002291 cites W3127419657 @default.
- W4313002291 cites W3173241699 @default.
- W4313002291 cites W3191739452 @default.
- W4313002291 cites W3203612130 @default.
- W4313002291 cites W4214897085 @default.
- W4313002291 cites W4230971985 @default.
- W4313002291 doi "https://doi.org/10.1007/978-3-031-19790-1_31" @default.
- W4313002291 hasPublicationYear "2022" @default.
- W4313002291 type Work @default.
- W4313002291 citedByCount "0" @default.
- W4313002291 crossrefType "book-chapter" @default.
- W4313002291 hasAuthorship W4313002291A5015956846 @default.
- W4313002291 hasAuthorship W4313002291A5026825380 @default.
- W4313002291 hasAuthorship W4313002291A5081710525 @default.
- W4313002291 hasAuthorship W4313002291A5085208414 @default.
- W4313002291 hasConcept C11413529 @default.
- W4313002291 hasConcept C115961682 @default.
- W4313002291 hasConcept C140745168 @default.
- W4313002291 hasConcept C153180895 @default.
- W4313002291 hasConcept C154945302 @default.
- W4313002291 hasConcept C160633673 @default.
- W4313002291 hasConcept C177264268 @default.
- W4313002291 hasConcept C199360897 @default.
- W4313002291 hasConcept C41008148 @default.
- W4313002291 hasConcept C80444323 @default.
- W4313002291 hasConceptScore W4313002291C11413529 @default.
- W4313002291 hasConceptScore W4313002291C115961682 @default.
- W4313002291 hasConceptScore W4313002291C140745168 @default.
- W4313002291 hasConceptScore W4313002291C153180895 @default.
- W4313002291 hasConceptScore W4313002291C154945302 @default.
- W4313002291 hasConceptScore W4313002291C160633673 @default.
- W4313002291 hasConceptScore W4313002291C177264268 @default.
- W4313002291 hasConceptScore W4313002291C199360897 @default.
- W4313002291 hasConceptScore W4313002291C41008148 @default.
- W4313002291 hasConceptScore W4313002291C80444323 @default.
- W4313002291 hasLocation W43130022911 @default.
- W4313002291 hasOpenAccess W4313002291 @default.
- W4313002291 hasPrimaryLocation W43130022911 @default.
- W4313002291 hasRelatedWork W1965781815 @default.
- W4313002291 hasRelatedWork W2136485282 @default.
- W4313002291 hasRelatedWork W2149249189 @default.
- W4313002291 hasRelatedWork W2269705005 @default.
- W4313002291 hasRelatedWork W2542880803 @default.
- W4313002291 hasRelatedWork W2546871836 @default.
- W4313002291 hasRelatedWork W2547748020 @default.
- W4313002291 hasRelatedWork W2918100922 @default.
- W4313002291 hasRelatedWork W3043252291 @default.
- W4313002291 hasRelatedWork W3209718219 @default.
- W4313002291 isParatext "false" @default.
- W4313002291 isRetracted "false" @default.
- W4313002291 workType "book-chapter" @default.