Anti-Spoofing and Mask-Supported Face Authentication Using mmWave Without On-Site Registration

Wenfan Song, Weiye Xu, Jianwei Liu, Yuanqing Zheng, Xinhuai Wang, Xian Xu, Jinsong Han

Abstract: Face authentication (FA) schemes are universally adopted. However, current FA systems are mainly camera-based and susceptible to masks and vulnerable to spoofing attacks. This paper exploits the penetrability, material sensitivity, and fine-grainedsensing capability of millimeter wave (mmWave) to build an anti-spoofing FA system, named mmFace. It scans faces by movinga commodity mmWave radar along a specific trajectory. The signals bounced off the face carry facial biometric and structure features, which allows mmFace to achieve reliable liveness detection and FA. Due to the penetrability of mmWave, mmFace can still work well when users wear masks. To en- hance security, we develop a liveness detection method and an amplitude modulation-based method to defend against spoofing attacks and replay attacks. We enhance the basic version of mmFace (Xu et al., 2022) by improving its performance under mask occlusion and replay attack resilience. Besides, we explore a distance-resistant structure feature to suppress the impact of unstable face- to-device distance. To avoid on-site registration, we propose a novel virtual registration approach based on the cross-modal transformation from photos to mmWave. We implement mmFace with various antenna configurations and prototype two typical modes of mmFace. Extensive experiments demonstrate mmFace’s accuracy in FA and effectiveness in attack detection.