Work

Extracting Local Information from Global Representations for Interpretable Deepfake Detection*