Matches in SemOpenAlex for { <https://semopenalex.org/work/W3153989314> ?p ?o ?g. }
- W3153989314 abstract "Document layout comprises both structural and visual (eg. font-sizes) information that is vital but often ignored by machine learning models. The few existing models which do use layout information only consider textual contents, and overlook the existence of contents in other modalities such as images. Additionally, spatial interactions of presented contents in a layout were never really fully exploited. To bridge this gap, we parse a document into content blocks (eg. text, table, image) and propose a novel layout-aware multimodal hierarchical framework, LAMPreT, to model the blocks and the whole document. Our LAMPreT encodes each block with a multimodal transformer in the lower-level and aggregates the block-level representations and connections utilizing a specifically designed transformer at the higher-level. We design hierarchical pretraining objectives where the lower-level model is trained similarly to multimodal grounding models, and the higher-level model is trained with our proposed novel layout-aware objectives. We evaluate the proposed model on two layout-aware tasks -- text block filling and image suggestion and show the effectiveness of our proposed hierarchical architecture as well as pretraining techniques." @default.
- W3153989314 created "2021-04-26" @default.
- W3153989314 creator A5024818901 @default.
- W3153989314 creator A5032248436 @default.
- W3153989314 creator A5035083907 @default.
- W3153989314 creator A5071466474 @default.
- W3153989314 creator A5080002047 @default.
- W3153989314 creator A5086723527 @default.
- W3153989314 date "2021-04-16" @default.
- W3153989314 modified "2023-09-28" @default.
- W3153989314 title "LAMPRET: Layout-Aware Multimodal PreTraining for Document Understanding" @default.
- W3153989314 cites W1544505229 @default.
- W3153989314 cites W1803802947 @default.
- W3153989314 cites W205625539 @default.
- W3153989314 cites W2138621090 @default.
- W3153989314 cites W2525778437 @default.
- W3153989314 cites W2613718673 @default.
- W3153989314 cites W2623808377 @default.
- W3153989314 cites W2891117443 @default.
- W3153989314 cites W2953384591 @default.
- W3153989314 cites W2963341956 @default.
- W3153989314 cites W2963403868 @default.
- W3153989314 cites W2964121744 @default.
- W3153989314 cites W2964346820 @default.
- W3153989314 cites W2968124245 @default.
- W3153989314 cites W2970042171 @default.
- W3153989314 cites W2970608575 @default.
- W3153989314 cites W2994673210 @default.
- W3153989314 cites W2999038738 @default.
- W3153989314 cites W3015468748 @default.
- W3153989314 cites W3017171774 @default.
- W3153989314 cites W3038476992 @default.
- W3153989314 cites W3040920819 @default.
- W3153989314 cites W3045733172 @default.
- W3153989314 cites W3090669478 @default.
- W3153989314 cites W3091806362 @default.
- W3153989314 cites W3093218477 @default.
- W3153989314 cites W3098440928 @default.
- W3153989314 cites W3098562592 @default.
- W3153989314 cites W3099103595 @default.
- W3153989314 cites W3104953317 @default.
- W3153989314 cites W3113753692 @default.
- W3153989314 cites W3188447078 @default.
- W3153989314 doi "https://doi.org/10.48550/arxiv.2104.08405" @default.
- W3153989314 hasPublicationYear "2021" @default.
- W3153989314 type Work @default.
- W3153989314 sameAs 3153989314 @default.
- W3153989314 citedByCount "2" @default.
- W3153989314 countsByYear W31539893142021 @default.
- W3153989314 countsByYear W31539893142022 @default.
- W3153989314 crossrefType "posted-content" @default.
- W3153989314 hasAuthorship W3153989314A5024818901 @default.
- W3153989314 hasAuthorship W3153989314A5032248436 @default.
- W3153989314 hasAuthorship W3153989314A5035083907 @default.
- W3153989314 hasAuthorship W3153989314A5071466474 @default.
- W3153989314 hasAuthorship W3153989314A5080002047 @default.
- W3153989314 hasAuthorship W3153989314A5086723527 @default.
- W3153989314 hasBestOaLocation W31539893141 @default.
- W3153989314 hasConcept C112698675 @default.
- W3153989314 hasConcept C115961682 @default.
- W3153989314 hasConcept C121332964 @default.
- W3153989314 hasConcept C124101348 @default.
- W3153989314 hasConcept C144133560 @default.
- W3153989314 hasConcept C144986985 @default.
- W3153989314 hasConcept C154945302 @default.
- W3153989314 hasConcept C165801399 @default.
- W3153989314 hasConcept C186644900 @default.
- W3153989314 hasConcept C188985296 @default.
- W3153989314 hasConcept C204321447 @default.
- W3153989314 hasConcept C23123220 @default.
- W3153989314 hasConcept C2524010 @default.
- W3153989314 hasConcept C2777210771 @default.
- W3153989314 hasConcept C33923547 @default.
- W3153989314 hasConcept C41008148 @default.
- W3153989314 hasConcept C62520636 @default.
- W3153989314 hasConcept C66322947 @default.
- W3153989314 hasConcept C72773152 @default.
- W3153989314 hasConceptScore W3153989314C112698675 @default.
- W3153989314 hasConceptScore W3153989314C115961682 @default.
- W3153989314 hasConceptScore W3153989314C121332964 @default.
- W3153989314 hasConceptScore W3153989314C124101348 @default.
- W3153989314 hasConceptScore W3153989314C144133560 @default.
- W3153989314 hasConceptScore W3153989314C144986985 @default.
- W3153989314 hasConceptScore W3153989314C154945302 @default.
- W3153989314 hasConceptScore W3153989314C165801399 @default.
- W3153989314 hasConceptScore W3153989314C186644900 @default.
- W3153989314 hasConceptScore W3153989314C188985296 @default.
- W3153989314 hasConceptScore W3153989314C204321447 @default.
- W3153989314 hasConceptScore W3153989314C23123220 @default.
- W3153989314 hasConceptScore W3153989314C2524010 @default.
- W3153989314 hasConceptScore W3153989314C2777210771 @default.
- W3153989314 hasConceptScore W3153989314C33923547 @default.
- W3153989314 hasConceptScore W3153989314C41008148 @default.
- W3153989314 hasConceptScore W3153989314C62520636 @default.
- W3153989314 hasConceptScore W3153989314C66322947 @default.
- W3153989314 hasConceptScore W3153989314C72773152 @default.
- W3153989314 hasLocation W31539893141 @default.
- W3153989314 hasOpenAccess W3153989314 @default.
- W3153989314 hasPrimaryLocation W31539893141 @default.
- W3153989314 hasRelatedWork W1552159754 @default.