Artifacts-Disentangled Adversarial Learning for Deepfake Detection | IEEE Journals & Magazine | IEEE Xplore