Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136203352> ?p ?o ?g. }
- W3136203352 endingPage "645" @default.
- W3136203352 startingPage "634" @default.
- W3136203352 abstract "Zero-shot Learning (ZSL) is a highly non-trivial task to generalize from seen to unseen classes. In this paper, we propose spherical zero-shot learning (SZSL) to address the major challenges in ZSL. By decoupling the similarity metric in the spherical embedding space into radius and angle, our SZSL can map classes to hyperspherical surfaces of different radiuses, which greatly increases its flexibility. Specifically, we introduce the spherical alignment on angles to spread classes as uniformly as possible to alleviate the hubness problem and simultaneously preserve the inter-class semantic structure to make the alignment more reasonable. We also introduce the spherical calibration with a minimum entropy based regularizer by adopting a larger radius for unseen classes than seen classes to reduce the prediction bias. Extensive experiments on five middle-scale benchmarks and large-scale ImageNet dataset demonstrate that the proposed approach consistently achieves superior performance for the traditional and generalized settings of ZSL." @default.
- W3136203352 created "2021-03-29" @default.
- W3136203352 creator A5026486701 @default.
- W3136203352 creator A5032171811 @default.
- W3136203352 creator A5081754197 @default.
- W3136203352 creator A5084889194 @default.
- W3136203352 date "2022-02-01" @default.
- W3136203352 modified "2023-10-14" @default.
- W3136203352 title "Spherical Zero-Shot Learning" @default.
- W3136203352 cites W1492420801 @default.
- W3136203352 cites W2017814585 @default.
- W3136203352 cites W2031489346 @default.
- W3136203352 cites W2070148066 @default.
- W3136203352 cites W2077071968 @default.
- W3136203352 cites W2092961325 @default.
- W3136203352 cites W2097117768 @default.
- W3136203352 cites W2098411764 @default.
- W3136203352 cites W2117539524 @default.
- W3136203352 cites W2128532956 @default.
- W3136203352 cites W2169153112 @default.
- W3136203352 cites W2171061940 @default.
- W3136203352 cites W2194775991 @default.
- W3136203352 cites W2289084343 @default.
- W3136203352 cites W2334493732 @default.
- W3136203352 cites W2400717490 @default.
- W3136203352 cites W2405223529 @default.
- W3136203352 cites W2552383788 @default.
- W3136203352 cites W2594833348 @default.
- W3136203352 cites W2611632661 @default.
- W3136203352 cites W2739770654 @default.
- W3136203352 cites W2789366140 @default.
- W3136203352 cites W2883124384 @default.
- W3136203352 cites W2887567284 @default.
- W3136203352 cites W2910453440 @default.
- W3136203352 cites W2915533878 @default.
- W3136203352 cites W2928966231 @default.
- W3136203352 cites W2948162522 @default.
- W3136203352 cites W2962716320 @default.
- W3136203352 cites W2962762077 @default.
- W3136203352 cites W2962903908 @default.
- W3136203352 cites W2963283377 @default.
- W3136203352 cites W2963466847 @default.
- W3136203352 cites W2963499153 @default.
- W3136203352 cites W2963538198 @default.
- W3136203352 cites W2963545832 @default.
- W3136203352 cites W2963846885 @default.
- W3136203352 cites W2963955422 @default.
- W3136203352 cites W2963960318 @default.
- W3136203352 cites W2963988212 @default.
- W3136203352 cites W2964086552 @default.
- W3136203352 cites W2964105864 @default.
- W3136203352 cites W2964162033 @default.
- W3136203352 cites W2972244714 @default.
- W3136203352 cites W2979402357 @default.
- W3136203352 cites W2979728922 @default.
- W3136203352 cites W3014645216 @default.
- W3136203352 cites W3034730995 @default.
- W3136203352 cites W3035297600 @default.
- W3136203352 cites W3035388827 @default.
- W3136203352 cites W3097309192 @default.
- W3136203352 cites W3100093508 @default.
- W3136203352 cites W3143107425 @default.
- W3136203352 cites W4232886157 @default.
- W3136203352 doi "https://doi.org/10.1109/tcsvt.2021.3067067" @default.
- W3136203352 hasPublicationYear "2022" @default.
- W3136203352 type Work @default.
- W3136203352 sameAs 3136203352 @default.
- W3136203352 citedByCount "8" @default.
- W3136203352 countsByYear W31362033522021 @default.
- W3136203352 countsByYear W31362033522023 @default.
- W3136203352 crossrefType "journal-article" @default.
- W3136203352 hasAuthorship W3136203352A5026486701 @default.
- W3136203352 hasAuthorship W3136203352A5032171811 @default.
- W3136203352 hasAuthorship W3136203352A5081754197 @default.
- W3136203352 hasAuthorship W3136203352A5084889194 @default.
- W3136203352 hasConcept C106301342 @default.
- W3136203352 hasConcept C11413529 @default.
- W3136203352 hasConcept C121332964 @default.
- W3136203352 hasConcept C127413603 @default.
- W3136203352 hasConcept C133731056 @default.
- W3136203352 hasConcept C138885662 @default.
- W3136203352 hasConcept C154945302 @default.
- W3136203352 hasConcept C162324750 @default.
- W3136203352 hasConcept C176217482 @default.
- W3136203352 hasConcept C205606062 @default.
- W3136203352 hasConcept C21547014 @default.
- W3136203352 hasConcept C2780813799 @default.
- W3136203352 hasConcept C33923547 @default.
- W3136203352 hasConcept C41008148 @default.
- W3136203352 hasConcept C41608201 @default.
- W3136203352 hasConcept C41895202 @default.
- W3136203352 hasConcept C62520636 @default.
- W3136203352 hasConceptScore W3136203352C106301342 @default.
- W3136203352 hasConceptScore W3136203352C11413529 @default.
- W3136203352 hasConceptScore W3136203352C121332964 @default.
- W3136203352 hasConceptScore W3136203352C127413603 @default.
- W3136203352 hasConceptScore W3136203352C133731056 @default.
- W3136203352 hasConceptScore W3136203352C138885662 @default.