Domain-generalizable Face Anti-Spoofing with Patch-based Multi-tasking and Artifact Pattern Conversion

TL;DR

This paper proposes PCGAN, a novel generative model that improves domain generalization in face anti-spoofing by disentangling artifacts and facial features, enhancing detection of unseen attacks.

Contribution

The paper introduces PCGAN, a patch-based multi-task GAN that generates diverse spoof artifacts and improves domain generalization in face anti-spoofing.

Findings

PCGAN significantly improves domain generalization in FAS.

The method effectively detects partial spoofing attacks.

Experiments show substantial security improvements in face recognition.

Abstract

Face Anti-Spoofing (FAS) algorithms, designed to secure face recognition systems against spoofing, struggle with limited dataset diversity, impairing their ability to handle unseen visual domains and spoofing methods. We introduce the Pattern Conversion Generative Adversarial Network (PCGAN) to enhance domain generalization in FAS. PCGAN effectively disentangles latent vectors for spoof artifacts and facial features, allowing to generate images with diverse artifacts. We further incorporate patch-based and multi-task learning to tackle partial attacks and overfitting issues to facial features. Our extensive experiments validate PCGAN's effectiveness in domain generalization and detecting partial attacks, giving a substantial improvement in facial recognition security.