Does this dataset contain object bounding box annotations?

[DiYaGo-zzh]

In Figure 1 of your paper, it appears to show bounding box annotations, but in the provided Annotations file, I couldn't find any related keys. The keys in the sequences dictionary are :
dict_keys(['width', 'height', 'id', 'seq_name', 'dataset', 'neg_category_ids', 'not_exhaustive_category_ids', 'fps', 'all_image_paths', 'annotated_image_paths', 'segmentations', 'track_category_ids']),
and there doesn't seem to be any bounding box labels. Does this dataset not include bounding box annotations?

[Ali2500]

The bounding boxes are not included in the JSON annotation files, but you can easily infer the box coordinates from the mask. With the toBbox() function in pycocotools this is possible even without decoding RLE into an image array.

[DiYaGo-zzh]

Thank you very much for your explanation!I now understand how to use the toBbox() function in pycocotools to directly infer bbox labels from the RLE format without decoding it into an image array. This has been incredibly helpful for our work—thank you again!