🤖 AI Summary
This study addresses the cross-national generalization challenge in identity document attack detection caused by the scarcity of real-world samples. We propose a few-shot/zero-shot detection framework based on prototypical networks and episodic training. Utilizing an EfficientNet-V2-b0 backbone, the method incorporates a four-sample prototypical head and a fixed-class variable-domain training mechanism to effectively learn universal attack cues and mitigate domain shift. Experiments demonstrate that the model achieves an average Equal Error Rate (EER) of approximately 9% across multi-national datasets and DLC-2021, significantly outperforming Softmax and CLIP baselines. These results indicate that the proposed approach provides an efficient and reliable solution for privacy-preserving cross-border remote account opening, effectively overcoming data limitations in diverse geographical contexts.
📝 Abstract
Presentation-Attack Detection (PAD) for national ID cards is limited by the lack of publicly available genuine samples, making it difficult for systems to generalize across countries. This paper introduces two main innovations: (1) a Prototypical Network head using an EfficientNet-V2-b0 backbone that requires only four genuine samples per class to create reliable prototypes; and (2) an episodic training regime that keeps PAD classes fixed while varying the card domain, allowing the network to learn universal attack cues.
Evaluated on a large multi-country dataset and the public DLC-2021 benchmark, this method achieves an average Equal Error Rate of around 9\%, outperforming conventional softmax and CLIP zero-shot baselines even with data from a single source country. This approach provides accurate, privacy-preserving PAD while minimizing data collection, facilitating scalable cross-jurisdictional remote onboarding.