Matches in SemOpenAlex for { <https://semopenalex.org/work/W4306802146> ?p ?o ?g. }
Showing items 1 to 75 of
75
with 100 items per page.
- W4306802146 abstract "Visual question answering (VQA) is a hallmark of vision and language reasoning and a challenging task under the zero-shot setting. We propose Plug-and-Play VQA (PNP-VQA), a modular framework for zero-shot VQA. In contrast to most existing works, which require substantial adaptation of pretrained language models (PLMs) for the vision modality, PNP-VQA requires no additional training of the PLMs. Instead, we propose to use natural language and network interpretation as an intermediate representation that glues pretrained models together. We first generate question-guided informative image captions, and pass the captions to a PLM as context for question answering. Surpassing end-to-end trained baselines, PNP-VQA achieves state-of-the-art results on zero-shot VQAv2 and GQA. With 11B parameters, it outperforms the 80B-parameter Flamingo model by 8.5% on VQAv2. With 738M PLM parameters, PNP-VQA achieves an improvement of 9.1% on GQA over FewVLM with 740M PLM parameters. Code is released at https://github.com/salesforce/LAVIS/tree/main/projects/pnp-vqa" @default.
- W4306802146 created "2022-10-20" @default.
- W4306802146 creator A5004557807 @default.
- W4306802146 creator A5037149973 @default.
- W4306802146 creator A5042646536 @default.
- W4306802146 creator A5065154024 @default.
- W4306802146 creator A5074834854 @default.
- W4306802146 date "2022-10-17" @default.
- W4306802146 modified "2023-10-18" @default.
- W4306802146 title "Plug-and-Play VQA: Zero-shot VQA by Conjoining Large Pretrained Models with Zero Training" @default.
- W4306802146 doi "https://doi.org/10.48550/arxiv.2210.08773" @default.
- W4306802146 hasPublicationYear "2022" @default.
- W4306802146 type Work @default.
- W4306802146 citedByCount "0" @default.
- W4306802146 crossrefType "posted-content" @default.
- W4306802146 hasAuthorship W4306802146A5004557807 @default.
- W4306802146 hasAuthorship W4306802146A5037149973 @default.
- W4306802146 hasAuthorship W4306802146A5042646536 @default.
- W4306802146 hasAuthorship W4306802146A5065154024 @default.
- W4306802146 hasAuthorship W4306802146A5074834854 @default.
- W4306802146 hasBestOaLocation W43068021461 @default.
- W4306802146 hasConcept C101468663 @default.
- W4306802146 hasConcept C119857082 @default.
- W4306802146 hasConcept C127413603 @default.
- W4306802146 hasConcept C138885662 @default.
- W4306802146 hasConcept C151730666 @default.
- W4306802146 hasConcept C154945302 @default.
- W4306802146 hasConcept C178790620 @default.
- W4306802146 hasConcept C185592680 @default.
- W4306802146 hasConcept C199360897 @default.
- W4306802146 hasConcept C201995342 @default.
- W4306802146 hasConcept C204321447 @default.
- W4306802146 hasConcept C2778344882 @default.
- W4306802146 hasConcept C2779343474 @default.
- W4306802146 hasConcept C2780451532 @default.
- W4306802146 hasConcept C2780813799 @default.
- W4306802146 hasConcept C41008148 @default.
- W4306802146 hasConcept C41895202 @default.
- W4306802146 hasConcept C44291984 @default.
- W4306802146 hasConcept C86803240 @default.
- W4306802146 hasConceptScore W4306802146C101468663 @default.
- W4306802146 hasConceptScore W4306802146C119857082 @default.
- W4306802146 hasConceptScore W4306802146C127413603 @default.
- W4306802146 hasConceptScore W4306802146C138885662 @default.
- W4306802146 hasConceptScore W4306802146C151730666 @default.
- W4306802146 hasConceptScore W4306802146C154945302 @default.
- W4306802146 hasConceptScore W4306802146C178790620 @default.
- W4306802146 hasConceptScore W4306802146C185592680 @default.
- W4306802146 hasConceptScore W4306802146C199360897 @default.
- W4306802146 hasConceptScore W4306802146C201995342 @default.
- W4306802146 hasConceptScore W4306802146C204321447 @default.
- W4306802146 hasConceptScore W4306802146C2778344882 @default.
- W4306802146 hasConceptScore W4306802146C2779343474 @default.
- W4306802146 hasConceptScore W4306802146C2780451532 @default.
- W4306802146 hasConceptScore W4306802146C2780813799 @default.
- W4306802146 hasConceptScore W4306802146C41008148 @default.
- W4306802146 hasConceptScore W4306802146C41895202 @default.
- W4306802146 hasConceptScore W4306802146C44291984 @default.
- W4306802146 hasConceptScore W4306802146C86803240 @default.
- W4306802146 hasLocation W43068021461 @default.
- W4306802146 hasOpenAccess W4306802146 @default.
- W4306802146 hasPrimaryLocation W43068021461 @default.
- W4306802146 hasRelatedWork W128392744 @default.
- W4306802146 hasRelatedWork W207304934 @default.
- W4306802146 hasRelatedWork W2081647779 @default.
- W4306802146 hasRelatedWork W2747680751 @default.
- W4306802146 hasRelatedWork W2809632469 @default.
- W4306802146 hasRelatedWork W2970044932 @default.
- W4306802146 hasRelatedWork W3107474891 @default.
- W4306802146 hasRelatedWork W3185852197 @default.
- W4306802146 hasRelatedWork W4297834192 @default.
- W4306802146 hasRelatedWork W4318620715 @default.
- W4306802146 isParatext "false" @default.
- W4306802146 isRetracted "false" @default.
- W4306802146 workType "article" @default.