Work

Cross-Domain GAN Removal: Evaluating Deepfake Detector Robustness via Artifact-Free Translation