CAOS is a framework that uses word-embedding similarities and an ensemble of vision-language models to detect and explain object hallucinations in image captioning.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Evaluating Hallucination in Large Vision-Language Models based on Context-Aware Object Similarities
CAOS is a framework that uses word-embedding similarities and an ensemble of vision-language models to detect and explain object hallucinations in image captioning.