MIG-COW: Transferable Adversarial Attacks on Deepfake Detectors via Gradient Decomposition
MIG-COW is proposed, a novel adversarial attack framework that generates highly generalizable and visually imperceptible adversarial examples that captures both shared and model-specific vulnerabilities across heterogeneous CNN and ViT detectors.
Wonju Seo, Joonhyuk Baek, Yeseong Jung et al.
· ACM Multimedia · 1 citation