A compact multimodal vision-text model for cross-domain ID card presentation attack detection generalizes well after fine-tuning but fails zero-shot, showing synthetic datasets may not capture real-world challenges.
Open-Set: ID Card Presentation Attack Detection Using Neural Style Transfer,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
From Vision to Text: A Compact Multimodal Approach for Robust, Cross-Domain Presentation Attack Detection on ID Cards
A compact multimodal vision-text model for cross-domain ID card presentation attack detection generalizes well after fine-tuning but fails zero-shot, showing synthetic datasets may not capture real-world challenges.