Yu-Ting Ta

Look Before You Judge: Training-Free Region Mining for Grounded and Explainable Deepfake Detection

Multimodal large language models (MLLMs) can explain deepfake verdicts in natural language, but such explanations are not necessarily visually grounded in the visual evidence underlying the prediction. A model may describe plausible artifacts inferred from language priors rather than from image evidence. Existing groun...