Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386562862> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W4386562862 abstract "Large-scale text-to-image diffusion models have shown impressive capabilities across various generative tasks, enabled by strong vision-language alignment obtained through pre-training. However, most vision-language discriminative tasks require extensive fine-tuning on carefully-labeled datasets to acquire such alignment, with great cost in time and computing resources. In this work, we explore directly applying a pre-trained generative diffusion model to the challenging discriminative task of visual grounding without any fine-tuning and additional training dataset. Specifically, we propose VGDiffZero, a simple yet effective zero-shot visual grounding framework based on text-to-image diffusion models. We also design a comprehensive region-scoring method considering both global and local contexts of each isolated proposal. Extensive experiments on RefCOCO, RefCOCO+, and RefCOCOg show that VGDiffZero achieves strong performance on zero-shot visual grounding." @default.
- W4386562862 created "2023-09-10" @default.
- W4386562862 creator A5012234902 @default.
- W4386562862 creator A5032326710 @default.
- W4386562862 creator A5034585885 @default.
- W4386562862 creator A5043876209 @default.
- W4386562862 creator A5082606612 @default.
- W4386562862 date "2023-09-03" @default.
- W4386562862 modified "2023-10-16" @default.
- W4386562862 title "VGDiffZero: Text-to-image Diffusion Models Can Be Zero-shot Visual Grounders" @default.
- W4386562862 doi "https://doi.org/10.48550/arxiv.2309.01141" @default.
- W4386562862 hasPublicationYear "2023" @default.
- W4386562862 type Work @default.
- W4386562862 citedByCount "0" @default.
- W4386562862 crossrefType "posted-content" @default.
- W4386562862 hasAuthorship W4386562862A5012234902 @default.
- W4386562862 hasAuthorship W4386562862A5032326710 @default.
- W4386562862 hasAuthorship W4386562862A5034585885 @default.
- W4386562862 hasAuthorship W4386562862A5043876209 @default.
- W4386562862 hasAuthorship W4386562862A5082606612 @default.
- W4386562862 hasBestOaLocation W43865628621 @default.
- W4386562862 hasConcept C115961682 @default.
- W4386562862 hasConcept C119857082 @default.
- W4386562862 hasConcept C121332964 @default.
- W4386562862 hasConcept C127413603 @default.
- W4386562862 hasConcept C138885662 @default.
- W4386562862 hasConcept C153180895 @default.
- W4386562862 hasConcept C154945302 @default.
- W4386562862 hasConcept C167966045 @default.
- W4386562862 hasConcept C178790620 @default.
- W4386562862 hasConcept C185592680 @default.
- W4386562862 hasConcept C201995342 @default.
- W4386562862 hasConcept C204321447 @default.
- W4386562862 hasConcept C2778344882 @default.
- W4386562862 hasConcept C2780451532 @default.
- W4386562862 hasConcept C2780813799 @default.
- W4386562862 hasConcept C31972630 @default.
- W4386562862 hasConcept C39890363 @default.
- W4386562862 hasConcept C41008148 @default.
- W4386562862 hasConcept C41895202 @default.
- W4386562862 hasConcept C69357855 @default.
- W4386562862 hasConcept C97355855 @default.
- W4386562862 hasConcept C97931131 @default.
- W4386562862 hasConceptScore W4386562862C115961682 @default.
- W4386562862 hasConceptScore W4386562862C119857082 @default.
- W4386562862 hasConceptScore W4386562862C121332964 @default.
- W4386562862 hasConceptScore W4386562862C127413603 @default.
- W4386562862 hasConceptScore W4386562862C138885662 @default.
- W4386562862 hasConceptScore W4386562862C153180895 @default.
- W4386562862 hasConceptScore W4386562862C154945302 @default.
- W4386562862 hasConceptScore W4386562862C167966045 @default.
- W4386562862 hasConceptScore W4386562862C178790620 @default.
- W4386562862 hasConceptScore W4386562862C185592680 @default.
- W4386562862 hasConceptScore W4386562862C201995342 @default.
- W4386562862 hasConceptScore W4386562862C204321447 @default.
- W4386562862 hasConceptScore W4386562862C2778344882 @default.
- W4386562862 hasConceptScore W4386562862C2780451532 @default.
- W4386562862 hasConceptScore W4386562862C2780813799 @default.
- W4386562862 hasConceptScore W4386562862C31972630 @default.
- W4386562862 hasConceptScore W4386562862C39890363 @default.
- W4386562862 hasConceptScore W4386562862C41008148 @default.
- W4386562862 hasConceptScore W4386562862C41895202 @default.
- W4386562862 hasConceptScore W4386562862C69357855 @default.
- W4386562862 hasConceptScore W4386562862C97355855 @default.
- W4386562862 hasConceptScore W4386562862C97931131 @default.
- W4386562862 hasLocation W43865628621 @default.
- W4386562862 hasOpenAccess W4386562862 @default.
- W4386562862 hasPrimaryLocation W43865628621 @default.
- W4386562862 hasRelatedWork W1576360539 @default.
- W4386562862 hasRelatedWork W2093104230 @default.
- W4386562862 hasRelatedWork W2770426046 @default.
- W4386562862 hasRelatedWork W2874782909 @default.
- W4386562862 hasRelatedWork W2888227225 @default.
- W4386562862 hasRelatedWork W2896673391 @default.
- W4386562862 hasRelatedWork W2952072295 @default.
- W4386562862 hasRelatedWork W2963402808 @default.
- W4386562862 hasRelatedWork W4289760695 @default.
- W4386562862 hasRelatedWork W2073139667 @default.
- W4386562862 isParatext "false" @default.
- W4386562862 isRetracted "false" @default.
- W4386562862 workType "article" @default.