Matches in SemOpenAlex for { <https://semopenalex.org/work/W4224931256> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W4224931256 abstract "Scene graph generation aims to describe the contents in scenes by identifying the objects and their relationships. In previous works, visual context is widely utilized in message passing networks to generate the representations for classification. However, the noisy estimation of visual context limits model performance. In this paper, we revisit the concept of incorporating visual context via a randomly ordered bidirectional Long Short Temporal Memory (biLSTM) based baseline, and show that noisy estimation is worse than random. To alleviate the problem, we propose a new method, dubbed Progressive Message Passing Network (PMP-Net) that better estimates the visual context in a coarse to fine manner. Specifically, we first estimate the visual context with a random initiated scene graph, then refine it with multi-head attention. The experimental results on the benchmark dataset Visual Genome show that PMP-Net achieves better or comparable performance on all three tasks: scene graph generation (SGGen), scene graph classification (SGCls), and predicate classification (PredCls)." @default.
- W4224931256 created "2022-04-28" @default.
- W4224931256 creator A5068837264 @default.
- W4224931256 creator A5070812231 @default.
- W4224931256 creator A5078084988 @default.
- W4224931256 creator A5090243913 @default.
- W4224931256 creator A5090930764 @default.
- W4224931256 date "2022-05-23" @default.
- W4224931256 modified "2023-09-27" @default.
- W4224931256 title "PMP-NET: Rethinking Visual Context for Scene Graph Generation" @default.
- W4224931256 cites W2077069816 @default.
- W4224931256 cites W2277195237 @default.
- W4224931256 cites W2579549467 @default.
- W4224931256 cites W2581887665 @default.
- W4224931256 cites W2607855566 @default.
- W4224931256 cites W2886970679 @default.
- W4224931256 cites W2962779575 @default.
- W4224931256 cites W2962785943 @default.
- W4224931256 cites W2963101956 @default.
- W4224931256 cites W2963536419 @default.
- W4224931256 cites W2963650529 @default.
- W4224931256 cites W2963938081 @default.
- W4224931256 cites W2969679616 @default.
- W4224931256 cites W3035017890 @default.
- W4224931256 cites W3106328333 @default.
- W4224931256 cites W3110636989 @default.
- W4224931256 cites W3181556077 @default.
- W4224931256 doi "https://doi.org/10.1109/icassp43922.2022.9747415" @default.
- W4224931256 hasPublicationYear "2022" @default.
- W4224931256 type Work @default.
- W4224931256 citedByCount "1" @default.
- W4224931256 countsByYear W42249312562023 @default.
- W4224931256 crossrefType "proceedings-article" @default.
- W4224931256 hasAuthorship W4224931256A5068837264 @default.
- W4224931256 hasAuthorship W4224931256A5070812231 @default.
- W4224931256 hasAuthorship W4224931256A5078084988 @default.
- W4224931256 hasAuthorship W4224931256A5090243913 @default.
- W4224931256 hasAuthorship W4224931256A5090930764 @default.
- W4224931256 hasConcept C119857082 @default.
- W4224931256 hasConcept C132525143 @default.
- W4224931256 hasConcept C13280743 @default.
- W4224931256 hasConcept C151730666 @default.
- W4224931256 hasConcept C153180895 @default.
- W4224931256 hasConcept C154945302 @default.
- W4224931256 hasConcept C179372163 @default.
- W4224931256 hasConcept C185798385 @default.
- W4224931256 hasConcept C205649164 @default.
- W4224931256 hasConcept C205711294 @default.
- W4224931256 hasConcept C2779343474 @default.
- W4224931256 hasConcept C36464697 @default.
- W4224931256 hasConcept C41008148 @default.
- W4224931256 hasConcept C80444323 @default.
- W4224931256 hasConcept C86803240 @default.
- W4224931256 hasConceptScore W4224931256C119857082 @default.
- W4224931256 hasConceptScore W4224931256C132525143 @default.
- W4224931256 hasConceptScore W4224931256C13280743 @default.
- W4224931256 hasConceptScore W4224931256C151730666 @default.
- W4224931256 hasConceptScore W4224931256C153180895 @default.
- W4224931256 hasConceptScore W4224931256C154945302 @default.
- W4224931256 hasConceptScore W4224931256C179372163 @default.
- W4224931256 hasConceptScore W4224931256C185798385 @default.
- W4224931256 hasConceptScore W4224931256C205649164 @default.
- W4224931256 hasConceptScore W4224931256C205711294 @default.
- W4224931256 hasConceptScore W4224931256C2779343474 @default.
- W4224931256 hasConceptScore W4224931256C36464697 @default.
- W4224931256 hasConceptScore W4224931256C41008148 @default.
- W4224931256 hasConceptScore W4224931256C80444323 @default.
- W4224931256 hasConceptScore W4224931256C86803240 @default.
- W4224931256 hasFunder F4320321001 @default.
- W4224931256 hasLocation W42249312561 @default.
- W4224931256 hasOpenAccess W4224931256 @default.
- W4224931256 hasPrimaryLocation W42249312561 @default.
- W4224931256 hasRelatedWork W12829028 @default.
- W4224931256 hasRelatedWork W13704299 @default.
- W4224931256 hasRelatedWork W2701911 @default.
- W4224931256 hasRelatedWork W306509 @default.
- W4224931256 hasRelatedWork W3979659 @default.
- W4224931256 hasRelatedWork W6266756 @default.
- W4224931256 hasRelatedWork W728297 @default.
- W4224931256 hasRelatedWork W7724241 @default.
- W4224931256 hasRelatedWork W8197146 @default.
- W4224931256 hasRelatedWork W9761094 @default.
- W4224931256 isParatext "false" @default.
- W4224931256 isRetracted "false" @default.
- W4224931256 workType "article" @default.