Domain Generalization with Pseudo-Domain Label for Face Anti-Spoofing

TL;DR

This paper introduces a novel face anti-spoofing method that automatically identifies pseudo-domains using feature statistics and depth estimators, improving generalization to unseen attack types.

Contribution

It proposes a domain generalization approach that self-assigns pseudo-domain labels via clustered feature statistics and depth information, without relying on dataset labels.

Findings

Improved detection of unseen face spoofing attacks.

Effective domain generalization demonstrated across multiple datasets.

Enhanced robustness without explicit domain labels.

Abstract

Face anti-spoofing (FAS) plays an important role in protecting face recognition systems from face representation attacks. Many recent studies in FAS have approached this problem with domain generalization technique. Domain generalization aims to increase generalization performance to better detect various types of attacks and unseen attacks. However, previous studies in this area have defined each domain simply as an anti-spoofing datasets and focused on developing learning techniques. In this paper, we proposed a method that enables network to judge its domain by itself with the clustered convolutional feature statistics from intermediate layers of the network, without labeling domains as datasets. We obtained pseudo-domain labels by not only using the network extracting features, but also using depth estimators, which were previously used only as an auxiliary task in FAS. In our experiments, we trained with three datasets and evaluated the performance with the remaining one dataset to demonstrate the effectiveness of the proposed method by conducting a total of four sets of experiments.