Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378465246> ?p ?o ?g. }
Showing items 1 to 63 of
63
with 100 items per page.
- W4378465246 abstract "Metrics for Visual Grounding (VG) in Visual Question Answering (VQA) systems primarily aim to measure a system's reliance on relevant parts of the image when inferring an answer to the given question. Lack of VG has been a common problem among state-of-the-art VQA systems and can manifest in over-reliance on irrelevant image parts or a disregard for the visual modality entirely. Although inference capabilities of VQA models are often illustrated by a few qualitative illustrations, most systems are not quantitatively assessed for their VG properties. We believe, an easily calculated criterion for meaningfully measuring a system's VG can help remedy this shortcoming, as well as add another valuable dimension to model evaluations and analysis. To this end, we propose a new VG metric that captures if a model a) identifies question-relevant objects in the scene, and b) actually relies on the information contained in the relevant objects when producing its answer, i.e., if its visual grounding is both faithful and plausible. Our metric, called Faithful and Plausible Visual Grounding (FPVG), is straightforward to determine for most VQA model designs. We give a detailed description of FPVG and evaluate several reference systems spanning various VQA architectures. Code to support the metric calculations on the GQA data set is available on GitHub." @default.
- W4378465246 created "2023-05-27" @default.
- W4378465246 creator A5010913778 @default.
- W4378465246 creator A5024904743 @default.
- W4378465246 creator A5058049725 @default.
- W4378465246 date "2023-05-24" @default.
- W4378465246 modified "2023-09-25" @default.
- W4378465246 title "Measuring Faithful and Plausible Visual Grounding in VQA" @default.
- W4378465246 doi "https://doi.org/10.48550/arxiv.2305.15015" @default.
- W4378465246 hasPublicationYear "2023" @default.
- W4378465246 type Work @default.
- W4378465246 citedByCount "0" @default.
- W4378465246 crossrefType "posted-content" @default.
- W4378465246 hasAuthorship W4378465246A5010913778 @default.
- W4378465246 hasAuthorship W4378465246A5024904743 @default.
- W4378465246 hasAuthorship W4378465246A5058049725 @default.
- W4378465246 hasBestOaLocation W43784652461 @default.
- W4378465246 hasConcept C124101348 @default.
- W4378465246 hasConcept C154945302 @default.
- W4378465246 hasConcept C162324750 @default.
- W4378465246 hasConcept C176217482 @default.
- W4378465246 hasConcept C177264268 @default.
- W4378465246 hasConcept C199360897 @default.
- W4378465246 hasConcept C202444582 @default.
- W4378465246 hasConcept C21547014 @default.
- W4378465246 hasConcept C23123220 @default.
- W4378465246 hasConcept C2776214188 @default.
- W4378465246 hasConcept C33676613 @default.
- W4378465246 hasConcept C33923547 @default.
- W4378465246 hasConcept C41008148 @default.
- W4378465246 hasConcept C44291984 @default.
- W4378465246 hasConcept C80444323 @default.
- W4378465246 hasConceptScore W4378465246C124101348 @default.
- W4378465246 hasConceptScore W4378465246C154945302 @default.
- W4378465246 hasConceptScore W4378465246C162324750 @default.
- W4378465246 hasConceptScore W4378465246C176217482 @default.
- W4378465246 hasConceptScore W4378465246C177264268 @default.
- W4378465246 hasConceptScore W4378465246C199360897 @default.
- W4378465246 hasConceptScore W4378465246C202444582 @default.
- W4378465246 hasConceptScore W4378465246C21547014 @default.
- W4378465246 hasConceptScore W4378465246C23123220 @default.
- W4378465246 hasConceptScore W4378465246C2776214188 @default.
- W4378465246 hasConceptScore W4378465246C33676613 @default.
- W4378465246 hasConceptScore W4378465246C33923547 @default.
- W4378465246 hasConceptScore W4378465246C41008148 @default.
- W4378465246 hasConceptScore W4378465246C44291984 @default.
- W4378465246 hasConceptScore W4378465246C80444323 @default.
- W4378465246 hasLocation W43784652461 @default.
- W4378465246 hasOpenAccess W4378465246 @default.
- W4378465246 hasPrimaryLocation W43784652461 @default.
- W4378465246 hasRelatedWork W1512698090 @default.
- W4378465246 hasRelatedWork W15319282 @default.
- W4378465246 hasRelatedWork W1602736231 @default.
- W4378465246 hasRelatedWork W2056129861 @default.
- W4378465246 hasRelatedWork W2118091901 @default.
- W4378465246 hasRelatedWork W2123793327 @default.
- W4378465246 hasRelatedWork W2135033253 @default.
- W4378465246 hasRelatedWork W2233955765 @default.
- W4378465246 hasRelatedWork W2366644548 @default.
- W4378465246 hasRelatedWork W71961486 @default.
- W4378465246 isParatext "false" @default.
- W4378465246 isRetracted "false" @default.
- W4378465246 workType "article" @default.