Interpretable-through-prototypes deepfake detection for diffusion models | IEEE Conference Publication | IEEE Xplore