Dual Defense: Adversarial, Traceable, and Invisible Robust Watermarking against Face Swapping

TL;DR

This paper introduces Dual Defense, a novel watermarking approach that invisibly embeds robust watermarks into facial images to actively prevent face swapping forgeries while maintaining traceability and high invisibility.

Contribution

The paper proposes a comprehensive active defense mechanism combining traceability and adversariality, with a new watermark embedding network based on feature impersonation attack.

Findings

Achieves high defense success rates against face swapping

Maintains watermark invisibility and robustness across datasets

Surpasses existing methods like CMUA-Watermark and FakeTagger

Abstract

The malicious applications of deep forgery, represented by face swapping, have introduced security threats such as misinformation dissemination and identity fraud. While some research has proposed the use of robust watermarking methods to trace the copyright of facial images for post-event traceability, these methods cannot effectively prevent the generation of forgeries at the source and curb their dissemination. To address this problem, we propose a novel comprehensive active defense mechanism that combines traceability and adversariality, called Dual Defense. Dual Defense invisibly embeds a single robust watermark within the target face to actively respond to sudden cases of malicious face swapping. It disrupts the output of the face swapping model while maintaining the integrity of watermark information throughout the entire dissemination process. This allows for watermark extraction at any stage of image tracking for traceability. Specifically, we introduce a watermark embedding network based on original-domain feature impersonation attack. This network learns robust adversarial features of target facial images and embeds watermarks, seeking a well-balanced trade-off between watermark invisibility, adversariality, and traceability through perceptual adversarial encoding strategies. Extensive experiments demonstrate that Dual Defense achieves optimal overall defense success rates and exhibits promising universality in anti-face swapping tasks and dataset generalization ability. It maintains impressive adversariality and traceability in both original and robust settings, surpassing current forgery defense methods that possess only one of these capabilities, including CMUA-Watermark, Anti-Forgery, FakeTagger, or PGD methods.