FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3167387588> ?p ?o ?g. }

• W3167387588 abstract "Image-level contrastive representation learning has proven to be highly effective as a generic model for transfer learning. Such generality for transfer learning, however, sacrifices specificity if we are interested in a certain downstream task. We argue that this could be sub-optimal and thus advocate a design principle which encourages alignment between the self-supervised pretext task and the downstream task. In this paper, we follow this principle with a pretraining method specifically designed for the task of object detection. We attain alignment in the following three aspects: 1) object-level representations are introduced via selective search bounding boxes as object proposals; 2) the pretraining network architecture incorporates the same dedicated modules used in the detection pipeline (e.g. FPN); 3) the pretraining is equipped with object detection properties such as object-level translation invariance and scale invariance. Our method, called Selective Object COntrastive learning (SoCo), achieves state-of-the-art results for transfer performance on COCO detection using a Mask R-CNN framework. Code is available at https://github.com/hologerry/SoCo." @default.
• W3167387588 created "2021-06-22" @default.
• W3167387588 creator A5001956754 @default.
• W3167387588 creator A5025114648 @default.
• W3167387588 creator A5059765313 @default.
• W3167387588 creator A5084464680 @default.
• W3167387588 creator A5090973869 @default.
• W3167387588 date "2021-06-04" @default.
• W3167387588 modified "2023-09-23" @default.
• W3167387588 title "Aligning Pretraining for Detection via Object-Level Contrastive Learning" @default.
• W3167387588 cites W1536680647 @default.
• W3167387588 cites W1665214252 @default.
• W3167387588 cites W1836465849 @default.
• W3167387588 cites W1861492603 @default.
• W3167387588 cites W189596042 @default.
• W3167387588 cites W1991367009 @default.
• W3167387588 cites W2031489346 @default.
• W3167387588 cites W2088049833 @default.
• W3167387588 cites W2102605133 @default.
• W3167387588 cites W2145094598 @default.
• W3167387588 cites W2194775991 @default.
• W3167387588 cites W2518108298 @default.
• W3167387588 cites W2565639579 @default.
• W3167387588 cites W2599837529 @default.
• W3167387588 cites W2622263826 @default.
• W3167387588 cites W2749988060 @default.
• W3167387588 cites W2757910899 @default.
• W3167387588 cites W2768489488 @default.
• W3167387588 cites W2798991696 @default.
• W3167387588 cites W2944828972 @default.
• W3167387588 cites W2948672349 @default.
• W3167387588 cites W2950141105 @default.
• W3167387588 cites W2953106684 @default.
• W3167387588 cites W2962742544 @default.
• W3167387588 cites W2963150697 @default.
• W3167387588 cites W2963351448 @default.
• W3167387588 cites W2963420272 @default.
• W3167387588 cites W2982770724 @default.
• W3167387588 cites W3009561768 @default.
• W3167387588 cites W3026092005 @default.
• W3167387588 cites W3034978746 @default.
• W3167387588 cites W3035058308 @default.
• W3167387588 cites W3035060554 @default.
• W3167387588 cites W3035524453 @default.
• W3167387588 cites W3036224891 @default.
• W3167387588 cites W3106528393 @default.
• W3167387588 cites W3106539090 @default.
• W3167387588 cites W3107668149 @default.
• W3167387588 cites W3134652006 @default.
• W3167387588 cites W3135715136 @default.
• W3167387588 cites W3135958856 @default.
• W3167387588 cites W3137901979 @default.
• W3167387588 cites W3159481202 @default.
• W3167387588 cites W3165924482 @default.
• W3167387588 cites W3172615411 @default.
• W3167387588 cites W3185102306 @default.
• W3167387588 cites W343636949 @default.
• W3167387588 cites W3108262825 @default.
• W3167387588 doi "https://doi.org/10.48550/arxiv.2106.02637" @default.
• W3167387588 hasPublicationYear "2021" @default.
• W3167387588 type Work @default.
• W3167387588 sameAs 3167387588 @default.
• W3167387588 citedByCount "1" @default.
• W3167387588 countsByYear W31673875882021 @default.
• W3167387588 crossrefType "posted-content" @default.
• W3167387588 hasAuthorship W3167387588A5001956754 @default.
• W3167387588 hasAuthorship W3167387588A5025114648 @default.
• W3167387588 hasAuthorship W3167387588A5059765313 @default.
• W3167387588 hasAuthorship W3167387588A5084464680 @default.
• W3167387588 hasAuthorship W3167387588A5090973869 @default.
• W3167387588 hasBestOaLocation W31673875881 @default.
• W3167387588 hasConcept C115961682 @default.
• W3167387588 hasConcept C119857082 @default.
• W3167387588 hasConcept C127413603 @default.
• W3167387588 hasConcept C147037132 @default.
• W3167387588 hasConcept C150899416 @default.
• W3167387588 hasConcept C153180895 @default.
• W3167387588 hasConcept C154945302 @default.
• W3167387588 hasConcept C15744967 @default.
• W3167387588 hasConcept C17744445 @default.
• W3167387588 hasConcept C199360897 @default.
• W3167387588 hasConcept C199539241 @default.
• W3167387588 hasConcept C201995342 @default.
• W3167387588 hasConcept C2776151529 @default.
• W3167387588 hasConcept C2779627259 @default.
• W3167387588 hasConcept C2780451532 @default.
• W3167387588 hasConcept C2780767217 @default.
• W3167387588 hasConcept C2781238097 @default.
• W3167387588 hasConcept C41008148 @default.
• W3167387588 hasConcept C43521106 @default.
• W3167387588 hasConcept C542102704 @default.
• W3167387588 hasConcept C63584917 @default.
• W3167387588 hasConcept C94124525 @default.
• W3167387588 hasConcept C94625758 @default.
• W3167387588 hasConceptScore W3167387588C115961682 @default.
• W3167387588 hasConceptScore W3167387588C119857082 @default.
• W3167387588 hasConceptScore W3167387588C127413603 @default.
• W3167387588 hasConceptScore W3167387588C147037132 @default.
• W3167387588 hasConceptScore W3167387588C150899416 @default.
• W3167387588 hasConceptScore W3167387588C153180895 @default.

• next