Tex-ViT: A Generalizable, Robust, Texture-based dual-branch cross-attention deepfake detector

ViTRe-assign community