GAVEL introduces a joint task, dataset, and benchmark for verifying, explaining, and localizing caption-image misalignments, with a supervised baseline that improves grounding and explanation metrics over strong closed-source models.
arXiv preprint arXiv:2412.19531 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GAVEL: Grounded Caption Error Verification and Localization
GAVEL introduces a joint task, dataset, and benchmark for verifying, explaining, and localizing caption-image misalignments, with a supervised baseline that improves grounding and explanation metrics over strong closed-source models.