Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313145013> ?p ?o ?g. }
- W4313145013 abstract "Visual grounding is a task to locate the target indicated by a natural language expression. Existing methods extend the generic object detection framework to this problem. They base the visual grounding on the features from pre-generated proposals or anchors, and fuse these features with the text embeddings to locate the target mentioned by the text. However, modeling the visual features from these predefined locations may fail to fully exploit the visual context and attribute information in the text query, which limits their performance. In this paper, we propose a transformer-based framework for accurate visual grounding by establishing text-conditioned discriminative features and performing multi-stage cross-modal reasoning. Specifically, we develop a visual-linguistic verification module to focus the visual features on regions relevant to the textual descriptions while suppressing the unrelated areas. A language-guided feature encoder is also devised to aggregate the visual contexts of the target object to improve the object's distinctiveness. To retrieve the target from the encoded visual features, we further propose a multi-stage cross-modal decoder to iteratively speculate on the correlations between the image and text for accurate target localization. Extensive experiments on five widely used datasets validate the efficacy of our proposed components and demonstrate state-of-the-art performance." @default.
- W4313145013 created "2023-01-06" @default.
- W4313145013 creator A5005377211 @default.
- W4313145013 creator A5015337144 @default.
- W4313145013 creator A5037488807 @default.
- W4313145013 creator A5047332568 @default.
- W4313145013 creator A5071037763 @default.
- W4313145013 creator A5083581319 @default.
- W4313145013 date "2022-06-01" @default.
- W4313145013 modified "2023-10-09" @default.
- W4313145013 title "Improving Visual Grounding with Visual-Linguistic Verification and Iterative Reasoning" @default.
- W4313145013 cites W1773149199 @default.
- W4313145013 cites W2006147162 @default.
- W4313145013 cites W2194775991 @default.
- W4313145013 cites W2251512949 @default.
- W4313145013 cites W2558535589 @default.
- W4313145013 cites W2606473278 @default.
- W4313145013 cites W2770129969 @default.
- W4313145013 cites W2904910963 @default.
- W4313145013 cites W2962764817 @default.
- W4313145013 cites W2962766617 @default.
- W4313145013 cites W2963109634 @default.
- W4313145013 cites W2963150697 @default.
- W4313145013 cites W2963735856 @default.
- W4313145013 cites W2964022527 @default.
- W4313145013 cites W2964345792 @default.
- W4313145013 cites W2984121207 @default.
- W4313145013 cites W2986701260 @default.
- W4313145013 cites W2986755220 @default.
- W4313145013 cites W2986803748 @default.
- W4313145013 cites W2987734933 @default.
- W4313145013 cites W3026458074 @default.
- W4313145013 cites W3034772468 @default.
- W4313145013 cites W3035268124 @default.
- W4313145013 cites W3107422826 @default.
- W4313145013 cites W3110435696 @default.
- W4313145013 cites W3173364567 @default.
- W4313145013 cites W3174004334 @default.
- W4313145013 cites W3204090293 @default.
- W4313145013 cites W4205412097 @default.
- W4313145013 cites W4214490042 @default.
- W4313145013 cites W639708223 @default.
- W4313145013 doi "https://doi.org/10.1109/cvpr52688.2022.00928" @default.
- W4313145013 hasPublicationYear "2022" @default.
- W4313145013 type Work @default.
- W4313145013 citedByCount "12" @default.
- W4313145013 countsByYear W43131450132022 @default.
- W4313145013 countsByYear W43131450132023 @default.
- W4313145013 crossrefType "proceedings-article" @default.
- W4313145013 hasAuthorship W4313145013A5005377211 @default.
- W4313145013 hasAuthorship W4313145013A5015337144 @default.
- W4313145013 hasAuthorship W4313145013A5037488807 @default.
- W4313145013 hasAuthorship W4313145013A5047332568 @default.
- W4313145013 hasAuthorship W4313145013A5071037763 @default.
- W4313145013 hasAuthorship W4313145013A5083581319 @default.
- W4313145013 hasBestOaLocation W43131450132 @default.
- W4313145013 hasConcept C151730666 @default.
- W4313145013 hasConcept C153180895 @default.
- W4313145013 hasConcept C154945302 @default.
- W4313145013 hasConcept C165696696 @default.
- W4313145013 hasConcept C195324797 @default.
- W4313145013 hasConcept C204321447 @default.
- W4313145013 hasConcept C2777508537 @default.
- W4313145013 hasConcept C2779343474 @default.
- W4313145013 hasConcept C2781238097 @default.
- W4313145013 hasConcept C36464697 @default.
- W4313145013 hasConcept C38652104 @default.
- W4313145013 hasConcept C41008148 @default.
- W4313145013 hasConcept C44291984 @default.
- W4313145013 hasConcept C86803240 @default.
- W4313145013 hasConcept C97931131 @default.
- W4313145013 hasConceptScore W4313145013C151730666 @default.
- W4313145013 hasConceptScore W4313145013C153180895 @default.
- W4313145013 hasConceptScore W4313145013C154945302 @default.
- W4313145013 hasConceptScore W4313145013C165696696 @default.
- W4313145013 hasConceptScore W4313145013C195324797 @default.
- W4313145013 hasConceptScore W4313145013C204321447 @default.
- W4313145013 hasConceptScore W4313145013C2777508537 @default.
- W4313145013 hasConceptScore W4313145013C2779343474 @default.
- W4313145013 hasConceptScore W4313145013C2781238097 @default.
- W4313145013 hasConceptScore W4313145013C36464697 @default.
- W4313145013 hasConceptScore W4313145013C38652104 @default.
- W4313145013 hasConceptScore W4313145013C41008148 @default.
- W4313145013 hasConceptScore W4313145013C44291984 @default.
- W4313145013 hasConceptScore W4313145013C86803240 @default.
- W4313145013 hasConceptScore W4313145013C97931131 @default.
- W4313145013 hasFunder F4320321001 @default.
- W4313145013 hasFunder F4320321543 @default.
- W4313145013 hasFunder F4320335777 @default.
- W4313145013 hasLocation W43131450131 @default.
- W4313145013 hasLocation W43131450132 @default.
- W4313145013 hasOpenAccess W4313145013 @default.
- W4313145013 hasPrimaryLocation W43131450131 @default.
- W4313145013 hasRelatedWork W1972656095 @default.
- W4313145013 hasRelatedWork W2024160000 @default.
- W4313145013 hasRelatedWork W2061273563 @default.
- W4313145013 hasRelatedWork W2285052147 @default.
- W4313145013 hasRelatedWork W2729514902 @default.
- W4313145013 hasRelatedWork W2743258233 @default.
- W4313145013 hasRelatedWork W2773500201 @default.