Hello, I would like to ask if the pre-annotation model supports batch pre-annotation of images or if it can only be done frame by frame.