Matches in SemOpenAlex for { <https://semopenalex.org/work/W3112156746> ?p ?o ?g. }
- W3112156746 abstract "Few-shot learning methods offer pre-training techniques optimized for easier later adaptation of the model to new classes (unseen during training) using one or a few examples. This adaptivity to unseen classes is especially important for many practical applications where the pre-trained label space cannot remain fixed for effective use and the model needs to be specialized to support new categories on the fly. One particularly interesting scenario, essentially overlooked by the few-shot literature, is Coarse-to-Fine Few-Shot (C2FS), where the training classes (e.g. animals) are of much `coarser granularity' than the target (test) classes (e.g. breeds). A very practical example of C2FS is when the target classes are sub-classes of the training classes. Intuitively, it is especially challenging as (both regular and few-shot) supervised pre-training tends to learn to ignore intra-class variability which is essential for separating sub-classes. In this paper, we introduce a novel 'Angular normalization' module that allows to effectively combine supervised and self-supervised contrastive pre-training to approach the proposed C2FS task, demonstrating significant gains in a broad study over multiple baselines and datasets. We hope that this work will help to pave the way for future research on this new, challenging, and very practical topic of C2FS classification." @default.
- W3112156746 created "2020-12-21" @default.
- W3112156746 creator A5020676344 @default.
- W3112156746 creator A5046199993 @default.
- W3112156746 creator A5050323834 @default.
- W3112156746 creator A5052325109 @default.
- W3112156746 creator A5056247353 @default.
- W3112156746 creator A5072571599 @default.
- W3112156746 creator A5075906727 @default.
- W3112156746 date "2020-12-07" @default.
- W3112156746 modified "2023-09-27" @default.
- W3112156746 title "Fine-grained Angular Contrastive Learning with Coarse Labels" @default.
- W3112156746 cites W2163605009 @default.
- W3112156746 cites W2321533354 @default.
- W3112156746 cites W2326925005 @default.
- W3112156746 cites W2601450892 @default.
- W3112156746 cites W2604763608 @default.
- W3112156746 cites W2742093937 @default.
- W3112156746 cites W2753160622 @default.
- W3112156746 cites W2772752821 @default.
- W3112156746 cites W2786928087 @default.
- W3112156746 cites W2842511635 @default.
- W3112156746 cites W2915604253 @default.
- W3112156746 cites W2943605315 @default.
- W3112156746 cites W2945390523 @default.
- W3112156746 cites W2948171095 @default.
- W3112156746 cites W2949650786 @default.
- W3112156746 cites W2949718784 @default.
- W3112156746 cites W2962723986 @default.
- W3112156746 cites W2962742544 @default.
- W3112156746 cites W2962987395 @default.
- W3112156746 cites W2963070905 @default.
- W3112156746 cites W2963263347 @default.
- W3112156746 cites W2963341924 @default.
- W3112156746 cites W2963350370 @default.
- W3112156746 cites W2963532621 @default.
- W3112156746 cites W2963741406 @default.
- W3112156746 cites W2964032613 @default.
- W3112156746 cites W2964105864 @default.
- W3112156746 cites W2964112702 @default.
- W3112156746 cites W2964249870 @default.
- W3112156746 cites W2965949912 @default.
- W3112156746 cites W2970941416 @default.
- W3112156746 cites W2971071159 @default.
- W3112156746 cites W2981707695 @default.
- W3112156746 cites W2983459436 @default.
- W3112156746 cites W2986604550 @default.
- W3112156746 cites W2988205463 @default.
- W3112156746 cites W2988501586 @default.
- W3112156746 cites W2989568391 @default.
- W3112156746 cites W2991559096 @default.
- W3112156746 cites W2994633389 @default.
- W3112156746 cites W3012255272 @default.
- W3112156746 cites W3034885317 @default.
- W3112156746 cites W3034978746 @default.
- W3112156746 cites W3035060554 @default.
- W3112156746 cites W3035524453 @default.
- W3112156746 cites W3036224891 @default.
- W3112156746 cites W3048399222 @default.
- W3112156746 cites W3108009835 @default.
- W3112156746 cites W3108975329 @default.
- W3112156746 cites W3109083691 @default.
- W3112156746 cites W3110608229 @default.
- W3112156746 cites W3118608800 @default.
- W3112156746 cites W343636949 @default.
- W3112156746 hasPublicationYear "2020" @default.
- W3112156746 type Work @default.
- W3112156746 sameAs 3112156746 @default.
- W3112156746 citedByCount "3" @default.
- W3112156746 countsByYear W31121567462021 @default.
- W3112156746 crossrefType "posted-content" @default.
- W3112156746 hasAuthorship W3112156746A5020676344 @default.
- W3112156746 hasAuthorship W3112156746A5046199993 @default.
- W3112156746 hasAuthorship W3112156746A5050323834 @default.
- W3112156746 hasAuthorship W3112156746A5052325109 @default.
- W3112156746 hasAuthorship W3112156746A5056247353 @default.
- W3112156746 hasAuthorship W3112156746A5072571599 @default.
- W3112156746 hasAuthorship W3112156746A5075906727 @default.
- W3112156746 hasConcept C111919701 @default.
- W3112156746 hasConcept C119857082 @default.
- W3112156746 hasConcept C120665830 @default.
- W3112156746 hasConcept C121332964 @default.
- W3112156746 hasConcept C136886441 @default.
- W3112156746 hasConcept C139807058 @default.
- W3112156746 hasConcept C144024400 @default.
- W3112156746 hasConcept C154945302 @default.
- W3112156746 hasConcept C162324750 @default.
- W3112156746 hasConcept C177774035 @default.
- W3112156746 hasConcept C187736073 @default.
- W3112156746 hasConcept C19165224 @default.
- W3112156746 hasConcept C2777212361 @default.
- W3112156746 hasConcept C2780451532 @default.
- W3112156746 hasConcept C41008148 @default.
- W3112156746 hasConceptScore W3112156746C111919701 @default.
- W3112156746 hasConceptScore W3112156746C119857082 @default.
- W3112156746 hasConceptScore W3112156746C120665830 @default.
- W3112156746 hasConceptScore W3112156746C121332964 @default.
- W3112156746 hasConceptScore W3112156746C136886441 @default.
- W3112156746 hasConceptScore W3112156746C139807058 @default.
- W3112156746 hasConceptScore W3112156746C144024400 @default.