Joint Physical-Digital Facial Attack Detection Via Simulating Spoofing Clues

Xianhua He,Dashuang Liang,Song Yang,Zhanlong Hao,Hui Ma,Binjie Mao,Xi Li,Yao Wang,Pengfei Yan,Ajian Liu

2024-04-12

Abstract:Face recognition systems are frequently subjected to a variety of physical and digital attacks of different types. Previous methods have achieved satisfactory performance in scenarios that address physical attacks and digital attacks, respectively. However, few methods are considered to integrate a model that simultaneously addresses both physical and digital attacks, implying the necessity to develop and maintain multiple models. To jointly detect physical and digital attacks within a single model, we propose an innovative approach that can adapt to any network architecture. Our approach mainly contains two types of data augmentation, which we call Simulated Physical Spoofing Clues augmentation (SPSC) and Simulated Digital Spoofing Clues augmentation (SDSC). SPSC and SDSC augment live samples into simulated attack samples by simulating spoofing clues of physical and digital attacks, respectively, which significantly improve the capability of the model to detect "unseen" attack types. Extensive experiments show that SPSC and SDSC can achieve state-of-the-art generalization in Protocols 2.1 and 2.2 of the UniAttackData dataset, respectively. Our method won first place in "Unified Physical-Digital Face Attack Detection" of the 5th Face Anti-spoofing Challenge@CVPR2024. Our final submission obtains 3.75% APCER, 0.93% BPCER, and 2.34% ACER, respectively. Our code is available at

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the ability to simultaneously detect physical attacks and digital attacks in facial recognition systems. Existing methods are usually only effective in dealing with a single type of attack, that is, either physical attacks (such as print attacks, replay attacks) or digital attacks (such as digital forgeries, adversarial attacks). This results in the need to maintain multiple different models to comprehensively judge the final results, which not only increases the demand for computing resources but also makes development and maintenance more complicated. Therefore, the paper proposes an innovative method, aiming to simultaneously detect physical attacks and digital attacks through a single model, thereby reducing the dependence on multiple models and improving the overall efficiency and generalization ability of the system.

Specifically, the paper proposes two data augmentation techniques: Simulated Physical Spoofing Cue Enhancement (SPSC) and Simulated Digital Spoofing Cue Enhancement (SDSC). These two techniques are respectively used to simulate the spoofing cues in physical attacks and digital attacks. By converting live samples into simulated attack samples, the ability of the model to detect "unseen" attack types is significantly improved. The experimental results of these methods on the UniAttackData dataset show that they can achieve state - of - the - art generalization performance, especially when dealing with "unseen" attack types in Protocols 2.1 and 2.2. In addition, this method won the first place in the "Unified Physical - Digital Facial Attack Detection" project in the 5th Facial Anti - Spoofing Challenge of CVPR in 2024, further verifying its effectiveness and superiority.