Matches in SemOpenAlex for { <https://semopenalex.org/work/W4377372012> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W4377372012 abstract "Visual spatial description (VSD) aims to generate texts that describe the spatial relations of the given objects within images. Existing VSD work merely models the 2D geometrical vision features, thus inevitably falling prey to the problem of skewed spatial understanding of target objects. In this work, we investigate the incorporation of 3D scene features for VSD. With an external 3D scene extractor, we obtain the 3D objects and scene features for input images, based on which we construct a target object-centered 3D spatial scene graph (Go3D-S2G), such that we model the spatial semantics of target objects within the holistic 3D scenes. Besides, we propose a scene subgraph selecting mechanism, sampling topologically-diverse subgraphs from Go3D-S2G, where the diverse local structure features are navigated to yield spatially-diversified text generation. Experimental results on two VSD datasets demonstrate that our framework outperforms the baselines significantly, especially improving on the cases with complex visual spatial relations. Meanwhile, our method can produce more spatially-diversified generation. Code is available at https://github.com/zhaoyucs/VSD." @default.
- W4377372012 created "2023-05-23" @default.
- W4377372012 creator A5003027292 @default.
- W4377372012 creator A5004953265 @default.
- W4377372012 creator A5019435679 @default.
- W4377372012 creator A5049531727 @default.
- W4377372012 creator A5053204257 @default.
- W4377372012 creator A5068047224 @default.
- W4377372012 creator A5089404640 @default.
- W4377372012 date "2023-05-19" @default.
- W4377372012 modified "2023-10-14" @default.
- W4377372012 title "Generating Visual Spatial Description via Holistic 3D Scene Understanding" @default.
- W4377372012 doi "https://doi.org/10.48550/arxiv.2305.11768" @default.
- W4377372012 hasPublicationYear "2023" @default.
- W4377372012 type Work @default.
- W4377372012 citedByCount "0" @default.
- W4377372012 crossrefType "posted-content" @default.
- W4377372012 hasAuthorship W4377372012A5003027292 @default.
- W4377372012 hasAuthorship W4377372012A5004953265 @default.
- W4377372012 hasAuthorship W4377372012A5019435679 @default.
- W4377372012 hasAuthorship W4377372012A5049531727 @default.
- W4377372012 hasAuthorship W4377372012A5053204257 @default.
- W4377372012 hasAuthorship W4377372012A5068047224 @default.
- W4377372012 hasAuthorship W4377372012A5089404640 @default.
- W4377372012 hasBestOaLocation W43773720121 @default.
- W4377372012 hasConcept C132525143 @default.
- W4377372012 hasConcept C153180895 @default.
- W4377372012 hasConcept C154945302 @default.
- W4377372012 hasConcept C179372163 @default.
- W4377372012 hasConcept C184337299 @default.
- W4377372012 hasConcept C199360897 @default.
- W4377372012 hasConcept C205711294 @default.
- W4377372012 hasConcept C27511587 @default.
- W4377372012 hasConcept C2780801425 @default.
- W4377372012 hasConcept C2781238097 @default.
- W4377372012 hasConcept C31972630 @default.
- W4377372012 hasConcept C41008148 @default.
- W4377372012 hasConcept C80444323 @default.
- W4377372012 hasConceptScore W4377372012C132525143 @default.
- W4377372012 hasConceptScore W4377372012C153180895 @default.
- W4377372012 hasConceptScore W4377372012C154945302 @default.
- W4377372012 hasConceptScore W4377372012C179372163 @default.
- W4377372012 hasConceptScore W4377372012C184337299 @default.
- W4377372012 hasConceptScore W4377372012C199360897 @default.
- W4377372012 hasConceptScore W4377372012C205711294 @default.
- W4377372012 hasConceptScore W4377372012C27511587 @default.
- W4377372012 hasConceptScore W4377372012C2780801425 @default.
- W4377372012 hasConceptScore W4377372012C2781238097 @default.
- W4377372012 hasConceptScore W4377372012C31972630 @default.
- W4377372012 hasConceptScore W4377372012C41008148 @default.
- W4377372012 hasConceptScore W4377372012C80444323 @default.
- W4377372012 hasLocation W43773720121 @default.
- W4377372012 hasOpenAccess W4377372012 @default.
- W4377372012 hasPrimaryLocation W43773720121 @default.
- W4377372012 hasRelatedWork W1837097281 @default.
- W4377372012 hasRelatedWork W1966410754 @default.
- W4377372012 hasRelatedWork W2007544051 @default.
- W4377372012 hasRelatedWork W2030539674 @default.
- W4377372012 hasRelatedWork W2095705906 @default.
- W4377372012 hasRelatedWork W2166024367 @default.
- W4377372012 hasRelatedWork W2325242284 @default.
- W4377372012 hasRelatedWork W2363840281 @default.
- W4377372012 hasRelatedWork W2789220062 @default.
- W4377372012 hasRelatedWork W2975200075 @default.
- W4377372012 isParatext "false" @default.
- W4377372012 isRetracted "false" @default.
- W4377372012 workType "article" @default.