CultureCLIP fine-tunes CLIP on synthetic twin image-caption pairs of visually similar, culturally distinct concepts, improving fine-grained cultural accuracy by up to 5.49% while preserving general performance.
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
-
CultureCLIP: Empowering CLIP with Cultural Awareness through Synthetic Images and Contextualized Captions
CultureCLIP fine-tunes CLIP on synthetic twin image-caption pairs of visually similar, culturally distinct concepts, improving fine-grained cultural accuracy by up to 5.49% while preserving general performance.