AdvCloak: Customized adversarial cloak for privacy protection

Xuannan Liu, Yaoyao Zhong, Xing Cui, Yuhang Zhang, Peipei Li, Weihong Deng

Abstract: Highlights•We propose AdvCloak, a GAN-based privacy framework that customizes class-wise masks.•A Two-stage training method addresses visual naturalness and feature generalization.•Extensive experiments show that AdvCloak achieves SOTA face privacy protection.