Matches in SemOpenAlex for { <https://semopenalex.org/work/W3145782600> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W3145782600 abstract "Locating semantically meaningful landmark points is a crucial component of a large number of computer vision pipelines. Because of the small number of available datasets with ground truth landmark annotations, it is essential to design robust unsupervised and semi-supervised methods for landmark detection. Many of the recent unsupervised learning methods rely on the equivariance properties of landmarks to synthetic image deformations. Our work focuses on such widely used methods and sheds light on its core problem, its inability to produce equivariant intermediate convolutional features. This finding leads us to formulate a two-step unsupervised approach that overcomes this challenge by first learning powerful pixel-based features that are then exploited in a second step within standard landmark detection approaches. Our methodology produces state-of-the-art results on several benchmarks such as the BBC Pose dataset, the Cat-head dataset and performs comparably in other situations." @default.
- W3145782600 created "2021-04-13" @default.
- W3145782600 creator A5041037976 @default.
- W3145782600 creator A5060952088 @default.
- W3145782600 creator A5075419354 @default.
- W3145782600 date "2022-08-21" @default.
- W3145782600 modified "2023-09-26" @default.
- W3145782600 title "Pretrained Equivariant Features Improve Unsupervised Landmark Discovery" @default.
- W3145782600 cites W1795776638 @default.
- W3145782600 cites W1834627138 @default.
- W3145782600 cites W1896424170 @default.
- W3145782600 cites W1976948919 @default.
- W3145782600 cites W1986656583 @default.
- W3145782600 cites W1990937109 @default.
- W3145782600 cites W1998294030 @default.
- W3145782600 cites W2011700682 @default.
- W3145782600 cites W2051928158 @default.
- W3145782600 cites W2082308025 @default.
- W3145782600 cites W2116277445 @default.
- W3145782600 cites W2135132101 @default.
- W3145782600 cites W2147334734 @default.
- W3145782600 cites W2163998463 @default.
- W3145782600 cites W2412002662 @default.
- W3145782600 cites W2604233003 @default.
- W3145782600 cites W2780034403 @default.
- W3145782600 cites W2807725536 @default.
- W3145782600 cites W2952074561 @default.
- W3145782600 cites W2962981304 @default.
- W3145782600 cites W2963419579 @default.
- W3145782600 cites W2963993755 @default.
- W3145782600 cites W2987936369 @default.
- W3145782600 cites W602397586 @default.
- W3145782600 doi "https://doi.org/10.1109/icpr56361.2022.9956547" @default.
- W3145782600 hasPublicationYear "2022" @default.
- W3145782600 type Work @default.
- W3145782600 sameAs 3145782600 @default.
- W3145782600 citedByCount "1" @default.
- W3145782600 countsByYear W31457826002021 @default.
- W3145782600 crossrefType "proceedings-article" @default.
- W3145782600 hasAuthorship W3145782600A5041037976 @default.
- W3145782600 hasAuthorship W3145782600A5060952088 @default.
- W3145782600 hasAuthorship W3145782600A5075419354 @default.
- W3145782600 hasBestOaLocation W31457826002 @default.
- W3145782600 hasConcept C119857082 @default.
- W3145782600 hasConcept C146849305 @default.
- W3145782600 hasConcept C153180895 @default.
- W3145782600 hasConcept C154945302 @default.
- W3145782600 hasConcept C2780297707 @default.
- W3145782600 hasConcept C31972630 @default.
- W3145782600 hasConcept C41008148 @default.
- W3145782600 hasConcept C8038995 @default.
- W3145782600 hasConceptScore W3145782600C119857082 @default.
- W3145782600 hasConceptScore W3145782600C146849305 @default.
- W3145782600 hasConceptScore W3145782600C153180895 @default.
- W3145782600 hasConceptScore W3145782600C154945302 @default.
- W3145782600 hasConceptScore W3145782600C2780297707 @default.
- W3145782600 hasConceptScore W3145782600C31972630 @default.
- W3145782600 hasConceptScore W3145782600C41008148 @default.
- W3145782600 hasConceptScore W3145782600C8038995 @default.
- W3145782600 hasLocation W31457826001 @default.
- W3145782600 hasLocation W31457826002 @default.
- W3145782600 hasOpenAccess W3145782600 @default.
- W3145782600 hasPrimaryLocation W31457826001 @default.
- W3145782600 hasRelatedWork W166366606 @default.
- W3145782600 hasRelatedWork W1990932233 @default.
- W3145782600 hasRelatedWork W2016546218 @default.
- W3145782600 hasRelatedWork W2098911910 @default.
- W3145782600 hasRelatedWork W2098980211 @default.
- W3145782600 hasRelatedWork W2352223314 @default.
- W3145782600 hasRelatedWork W2509104183 @default.
- W3145782600 hasRelatedWork W2509618504 @default.
- W3145782600 hasRelatedWork W2525521397 @default.
- W3145782600 hasRelatedWork W2156243485 @default.
- W3145782600 isParatext "false" @default.
- W3145782600 isRetracted "false" @default.
- W3145782600 magId "3145782600" @default.
- W3145782600 workType "article" @default.