Matches in SemOpenAlex for { <https://semopenalex.org/work/W2955014993> ?p ?o ?g. }
- W2955014993 abstract "We focus on grounding (i.e., localizing or linking) referring expressions in images, e.g., ``largest elephant standing behind baby elephant''. This is a general yet challenging vision-language task since it does not only require the localization of objects, but also the multimodal comprehension of context -- visual attributes (e.g., ``largest'', ``baby'') and relationships (e.g., ``behind'') that help to distinguish the referent from other objects, especially those of the same category. Due to the exponential complexity involved in modeling the context associated with multiple image regions, existing work oversimplifies this task to pairwise region modeling by multiple instance learning. In this paper, we propose a variational Bayesian method, called Variational Context, to solve the problem of complex context modeling in referring expression grounding. Specifically, our framework exploits the reciprocal relation between the referent and context, i.e., either of them influences estimation of the posterior distribution of the other, and thereby the search space of context can be greatly reduced. In addition to reciprocity, our framework considers the semantic information of context, i.e., the referring expression can be reproduced based on the estimated context. We also extend the model to unsupervised setting where no annotation for the referent is available. Extensive experiments on various benchmarks show consistent improvement over state-of-the-art methods in both supervised and unsupervised settings." @default.
- W2955014993 created "2019-07-12" @default.
- W2955014993 creator A5028589167 @default.
- W2955014993 creator A5037340457 @default.
- W2955014993 creator A5042324027 @default.
- W2955014993 creator A5085349794 @default.
- W2955014993 date "2020-01-01" @default.
- W2955014993 modified "2023-10-16" @default.
- W2955014993 title "Variational Context: Exploiting Visual and Textual Context for Grounding Referring Expressions" @default.
- W2955014993 cites W1514535095 @default.
- W2955014993 cites W1533861849 @default.
- W2955014993 cites W1533917153 @default.
- W2955014993 cites W1543294385 @default.
- W2955014993 cites W15689187 @default.
- W2955014993 cites W1773149199 @default.
- W2955014993 cites W1836465849 @default.
- W2955014993 cites W1861492603 @default.
- W2955014993 cites W1895577753 @default.
- W2955014993 cites W1933349210 @default.
- W2955014993 cites W1959608418 @default.
- W2955014993 cites W2005814556 @default.
- W2955014993 cites W2110119381 @default.
- W2955014993 cites W2119717200 @default.
- W2955014993 cites W2125447031 @default.
- W2955014993 cites W2128248292 @default.
- W2955014993 cites W2131774270 @default.
- W2955014993 cites W2136634080 @default.
- W2955014993 cites W2156718681 @default.
- W2955014993 cites W2188365844 @default.
- W2955014993 cites W2250378130 @default.
- W2955014993 cites W2250539671 @default.
- W2955014993 cites W2250799982 @default.
- W2955014993 cites W2251512949 @default.
- W2955014993 cites W2479423890 @default.
- W2955014993 cites W2489434015 @default.
- W2955014993 cites W2554334383 @default.
- W2955014993 cites W2558535589 @default.
- W2955014993 cites W2570343428 @default.
- W2955014993 cites W2571175805 @default.
- W2955014993 cites W2583360688 @default.
- W2955014993 cites W2591644541 @default.
- W2955014993 cites W2600144439 @default.
- W2955014993 cites W2603266952 @default.
- W2955014993 cites W2605736949 @default.
- W2955014993 cites W2607855566 @default.
- W2955014993 cites W2609468337 @default.
- W2955014993 cites W2613718673 @default.
- W2955014993 cites W2623594919 @default.
- W2955014993 cites W2741903908 @default.
- W2955014993 cites W2770129969 @default.
- W2955014993 cites W2779827764 @default.
- W2955014993 cites W2799263800 @default.
- W2955014993 cites W2904164128 @default.
- W2955014993 cites W2912274075 @default.
- W2955014993 cites W2950761309 @default.
- W2955014993 cites W2962764817 @default.
- W2955014993 cites W2963042258 @default.
- W2955014993 cites W2963109634 @default.
- W2955014993 cites W2963224792 @default.
- W2955014993 cites W2963547393 @default.
- W2955014993 cites W2963567641 @default.
- W2955014993 cites W2963668159 @default.
- W2955014993 cites W2963735856 @default.
- W2955014993 cites W2963976294 @default.
- W2955014993 cites W2963980128 @default.
- W2955014993 cites W2964036520 @default.
- W2955014993 cites W2964199361 @default.
- W2955014993 cites W2964284374 @default.
- W2955014993 cites W2964308564 @default.
- W2955014993 cites W2964345792 @default.
- W2955014993 cites W3098232790 @default.
- W2955014993 cites W3106250896 @default.
- W2955014993 doi "https://doi.org/10.1109/tpami.2019.2926266" @default.
- W2955014993 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/31283493" @default.
- W2955014993 hasPublicationYear "2020" @default.
- W2955014993 type Work @default.
- W2955014993 sameAs 2955014993 @default.
- W2955014993 citedByCount "3" @default.
- W2955014993 countsByYear W29550149932020 @default.
- W2955014993 countsByYear W29550149932021 @default.
- W2955014993 crossrefType "journal-article" @default.
- W2955014993 hasAuthorship W2955014993A5028589167 @default.
- W2955014993 hasAuthorship W2955014993A5037340457 @default.
- W2955014993 hasAuthorship W2955014993A5042324027 @default.
- W2955014993 hasAuthorship W2955014993A5085349794 @default.
- W2955014993 hasBestOaLocation W29550149932 @default.
- W2955014993 hasConcept C119857082 @default.
- W2955014993 hasConcept C138885662 @default.
- W2955014993 hasConcept C151730666 @default.
- W2955014993 hasConcept C153180895 @default.
- W2955014993 hasConcept C154945302 @default.
- W2955014993 hasConcept C183322885 @default.
- W2955014993 hasConcept C184898388 @default.
- W2955014993 hasConcept C204321447 @default.
- W2955014993 hasConcept C2777096784 @default.
- W2955014993 hasConcept C2779343474 @default.
- W2955014993 hasConcept C2781238097 @default.
- W2955014993 hasConcept C41008148 @default.
- W2955014993 hasConcept C41895202 @default.
- W2955014993 hasConcept C64754055 @default.