It renders the input question on the image and predicts the answer.