Confusion about fixed_image_standardization

[njuaplusplus]

The following is the function fixed_image_standardization:

def fixed_image_standardization(image_tensor):
    processed_tensor = (image_tensor - 127.5) / 128.0
    return processed_tensor

fixed_image_standardization seems to normalize images in the range [0., 255.] into the range [-1.0, 1.0].

But the extract_face returns images in the range [0.0, 1.0].

Could anyone explain this? Is this normalization wrong?

Thank you.