发表机构
Facephi; Hochschule Darmstadt（菲思费（Facephi）; 达姆施塔特应用科学大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对身份证呈现攻击检测的跨国泛化难题，提出基于EfficientNet-V2-b0骨干的原型网络头与情节式训练机制，在跨国数据集及DLC-2021基准上实现约9%的平均等错误率，优于相关基线方法，可助力跨辖区远程入职。
AI 中文摘要
针对国家身份证的呈现攻击检测（PAD）受限于公开真实样本的缺失，导致系统难以跨国家泛化。本文提出两项主要创新：（1）采用EfficientNet-V2-b0作为骨干的原型网络头，每类仅需4个真实样本即可构建可靠原型；（2）采用情节式训练机制，固定PAD类别同时变更卡片域，使网络学习通用攻击线索。在大型跨国数据集及公开DLC-2021基准上评估，该方法平均等错误率约9%，即便仅使用单一来源国家的数据，也优于传统softmax及CLIP零样本基线。此方法提供准确、隐私保护的PAD，同时最小化数据收集，助力可扩展的跨辖区远程入职流程。
英文摘要
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.
CommentsPreprint accepted DAS 2026 at ICDAR 2026
