PVASS-MDD: Predictive Visual-Audio Alignment Self-Supervision for Multimodal Deepfake Detection | IEEE Journals & Magazine | IEEE Xplore