Matches in SemOpenAlex for { <https://semopenalex.org/work/W3174865181> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W3174865181 abstract "Traditional scene graph generation methods are trained using cross-entropy losses that treat objects and relationships as independent entities. Such a formulation, however, ignores the structure in the output space, in an inherently structured prediction problem. In this work, we introduce a novel energy-based learning framework for generating scene graphs. The proposed formulation allows for efficiently incorporating the structure of scene graphs in the output space. This additional constraint in the learning framework acts as an inductive bias and allows models to learn efficiently from a small number of labels. We use the proposed energy-based framework <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>†</sup> to train existing stateof-the-art models and obtain a significant performance improvement, of up to 21% and 27%, on the Visual Genome [9] and GQA [5] benchmark datasets, respectively. Furthermore, we showcase the learning efficiency of the proposed framework by demonstrating superior performance in the zero- and few-shot settings where data is scarce." @default.
- W3174865181 created "2021-07-05" @default.
- W3174865181 creator A5000243830 @default.
- W3174865181 creator A5006918685 @default.
- W3174865181 creator A5040593678 @default.
- W3174865181 creator A5049953837 @default.
- W3174865181 creator A5053011888 @default.
- W3174865181 creator A5070885109 @default.
- W3174865181 creator A5082081818 @default.
- W3174865181 date "2021-06-01" @default.
- W3174865181 modified "2023-10-18" @default.
- W3174865181 title "Energy-Based Learning for Scene Graph Generation" @default.
- W3174865181 cites W2277195237 @default.
- W3174865181 cites W2549139847 @default.
- W3174865181 cites W2579549467 @default.
- W3174865181 cites W2963101956 @default.
- W3174865181 cites W2963184176 @default.
- W3174865181 cites W2963518342 @default.
- W3174865181 cites W2963536419 @default.
- W3174865181 cites W2963938081 @default.
- W3174865181 cites W2992478697 @default.
- W3174865181 cites W3034538190 @default.
- W3174865181 cites W3035017890 @default.
- W3174865181 cites W3106759358 @default.
- W3174865181 doi "https://doi.org/10.1109/cvpr46437.2021.01372" @default.
- W3174865181 hasPublicationYear "2021" @default.
- W3174865181 type Work @default.
- W3174865181 sameAs 3174865181 @default.
- W3174865181 citedByCount "56" @default.
- W3174865181 countsByYear W31748651812021 @default.
- W3174865181 countsByYear W31748651812022 @default.
- W3174865181 countsByYear W31748651812023 @default.
- W3174865181 crossrefType "proceedings-article" @default.
- W3174865181 hasAuthorship W3174865181A5000243830 @default.
- W3174865181 hasAuthorship W3174865181A5006918685 @default.
- W3174865181 hasAuthorship W3174865181A5040593678 @default.
- W3174865181 hasAuthorship W3174865181A5049953837 @default.
- W3174865181 hasAuthorship W3174865181A5053011888 @default.
- W3174865181 hasAuthorship W3174865181A5070885109 @default.
- W3174865181 hasAuthorship W3174865181A5082081818 @default.
- W3174865181 hasBestOaLocation W31748651812 @default.
- W3174865181 hasConcept C106301342 @default.
- W3174865181 hasConcept C119857082 @default.
- W3174865181 hasConcept C121332964 @default.
- W3174865181 hasConcept C132525143 @default.
- W3174865181 hasConcept C13280743 @default.
- W3174865181 hasConcept C154945302 @default.
- W3174865181 hasConcept C185798385 @default.
- W3174865181 hasConcept C205649164 @default.
- W3174865181 hasConcept C36464697 @default.
- W3174865181 hasConcept C41008148 @default.
- W3174865181 hasConcept C62520636 @default.
- W3174865181 hasConcept C80444323 @default.
- W3174865181 hasConceptScore W3174865181C106301342 @default.
- W3174865181 hasConceptScore W3174865181C119857082 @default.
- W3174865181 hasConceptScore W3174865181C121332964 @default.
- W3174865181 hasConceptScore W3174865181C132525143 @default.
- W3174865181 hasConceptScore W3174865181C13280743 @default.
- W3174865181 hasConceptScore W3174865181C154945302 @default.
- W3174865181 hasConceptScore W3174865181C185798385 @default.
- W3174865181 hasConceptScore W3174865181C205649164 @default.
- W3174865181 hasConceptScore W3174865181C36464697 @default.
- W3174865181 hasConceptScore W3174865181C41008148 @default.
- W3174865181 hasConceptScore W3174865181C62520636 @default.
- W3174865181 hasConceptScore W3174865181C80444323 @default.
- W3174865181 hasLocation W31748651811 @default.
- W3174865181 hasLocation W31748651812 @default.
- W3174865181 hasOpenAccess W3174865181 @default.
- W3174865181 hasPrimaryLocation W31748651811 @default.
- W3174865181 hasRelatedWork W112744582 @default.
- W3174865181 hasRelatedWork W1485630101 @default.
- W3174865181 hasRelatedWork W1666765134 @default.
- W3174865181 hasRelatedWork W2498017833 @default.
- W3174865181 hasRelatedWork W2961085424 @default.
- W3174865181 hasRelatedWork W3081841992 @default.
- W3174865181 hasRelatedWork W4301605664 @default.
- W3174865181 hasRelatedWork W4301846872 @default.
- W3174865181 hasRelatedWork W4306674287 @default.
- W3174865181 hasRelatedWork W4224009465 @default.
- W3174865181 isParatext "false" @default.
- W3174865181 isRetracted "false" @default.
- W3174865181 magId "3174865181" @default.
- W3174865181 workType "article" @default.