Exposing Fine-Grained Adversarial Vulnerability of Face Anti-Spoofing Models

TL;DR

This paper introduces a framework to analyze and expose the fine-grained adversarial vulnerabilities of face anti-spoofing models, revealing specific weaknesses and aiding in robustness improvement.

Contribution

The paper proposes a novel framework with semantic feature augmentation to identify fine-grained adversarial vulnerabilities in face anti-spoofing models, enhancing analysis precision.

Findings

SFA module increases attack success rate by nearly 40%

Fine-grained analysis reveals model and feature vulnerabilities

Adversarial examples aid in selecting robust models and training

Abstract

Face anti-spoofing aims to discriminate the spoofing face images (e.g., printed photos) from live ones. However, adversarial examples greatly challenge its credibility, where adding some perturbation noise can easily change the predictions. Previous works conducted adversarial attack methods to evaluate the face anti-spoofing performance without any fine-grained analysis that which model architecture or auxiliary feature is vulnerable to the adversary. To handle this problem, we propose a novel framework to expose the fine-grained adversarial vulnerability of the face anti-spoofing models, which consists of a multitask module and a semantic feature augmentation (SFA) module. The multitask module can obtain different semantic features for further evaluation, but only attacking these semantic features fails to reflect the discrimination-related vulnerability. We then design the SFA module to introduce the data distribution prior for more discrimination-related gradient directions for generating adversarial examples. Comprehensive experiments show that SFA module increases the attack success rate by nearly 40$\%$ on average. We conduct this fine-grained adversarial analysis on different annotations, geometric maps, and backbone networks (e.g., Resnet network). These fine-grained adversarial examples can be used for selecting robust backbone networks and auxiliary features. They also can be used for adversarial training, which makes it practical to further improve the accuracy and robustness of the face anti-spoofing models.