# RoPAD: Robust Presentation Attack Detection through Unsupervised Adversarial Invariance

## TL;DR

RoPAD is a deep learning model that enhances presentation attack detection by using unsupervised adversarial invariance to ignore distractors, achieving state-of-the-art results across multiple benchmarks.

## Contribution

The paper introduces RoPAD, a novel end-to-end deep learning framework that employs unsupervised adversarial invariance to improve robustness against presentation attacks.

## Key findings

- Achieves state-of-the-art performance on benchmark datasets.
- Effectively ignores visual distractors in images.
- Reduces overfitting in presentation attack detection.

## Abstract

For enterprise, personal and societal applications, there is now an increasing demand for automated authentication of identity from images using computer vision. However, current authentication technologies are still vulnerable to presentation attacks. We present RoPAD, an end-to-end deep learning model for presentation attack detection that employs unsupervised adversarial invariance to ignore visual distractors in images for increased robustness and reduced overfitting. Experiments show that the proposed framework exhibits state-of-the-art performance on presentation attack detection on several benchmark datasets.