# Towards Zero-Shot Domain Generalization for ID Cards Presentation Attack Detection

arXiv: https://arxiv.org/abs/2608.16591 · HTML: https://mlsanity.com/paper/2608.16591/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 14 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Builds on (most cited first, 11 of 11)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [Learning to Compare: Relation Network for Few-Shot Learning](https://mlsanity.com/paper/1711.06025/) (2017)
- [EfficientNetV2: Smaller Models and Faster Training](https://mlsanity.com/paper/2104.00298/) (2021)
- [A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities](https://mlsanity.com/paper/2205.06743/) (2022)
- [MIDV-500: A Dataset for Identity Documents Analysis and Recognition on Mobile Devices in Video Stream](https://mlsanity.com/paper/1807.05786/) (2018)
- [MIDV-2020: A Comprehensive Benchmark Dataset for Identity Document Analysis](https://mlsanity.com/paper/2107.00396/) (2021)
- [First Competition on Presentation Attack Detection on ID Card](https://mlsanity.com/paper/2409.00372/) (2024)
- [Second Competition on Presentation Attack Detection on ID Card](https://mlsanity.com/paper/2507.20404/) (2025)
- [Few-Shot Learning: Expanding ID Cards Presentation Attack Detection to Unknown ID Countries](https://mlsanity.com/paper/2409.06842/) (2024)

## Abstract (arXiv)

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.