Below, we show a random sample of the true images from each of the dataset that we consider in our evaluation.