Question: How are the expressions sampled in object detection training?

[LingyvKong]

Hello Bin,
I trained on an object detection dataset with only 6 categories, but I found that expressions_new still does not contain all categories, what kind of sampling rules are there? I locate to func convert_object_detection_to_grounding_optimized_for_od in code /workspace/UNINEXT/projects/UNINEXT/uninext/data/coco_dataset_mapper_uni.py, I guess it means that all positive categories are kept and some negative categories are randomly selected? Why do I need to do this？

Thanks very much!