LightD creates natural adversarial relighting images with GPT-selected lighting parameters and gradient optimization, outperforming prior non-suspicious attacks on vision-language models.
APA measures the percentage of successful prediction answers among all the images
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
-
Light as Deception: GPT-driven Natural Relighting Against Vision-Language Pre-training Models
LightD creates natural adversarial relighting images with GPT-selected lighting parameters and gradient optimization, outperforming prior non-suspicious attacks on vision-language models.