VORD filters hallucinated tokens by requiring that a token's probability on the original image exceed its probability on a corrupted image, with an adaptive margin.
Towards causal vqa: Revealing and reducing spurious correlations by invariant and covariant semantic editing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
VORD: Visual Ordinal Calibration for Mitigating Object Hallucinations in Large Vision-Language Models
VORD filters hallucinated tokens by requiring that a token's probability on the original image exceed its probability on a corrupted image, with an adaptive margin.