bbox AP not being calculated

[Changomango0903]

I've been trying to evaluate my YOLO models using this tool but I can't seem to get the AP to calculate. Attached below are the coco-labels and the results.json that I've created matching the example mask-rcnn format that was shown in the jupyter notebook. data.json _annotations.coco.json

[Changomango0903]

Incase it is also required, I will include the code used to evaluate the results below:

bbox_file = 'path to json'

from tidecv import TIDE
import tidecv.datasets as datasets

dataset = datasets.COCO("path-to-annotation")
bbox_results = datasets.COCOResult(bbox_file)
tide = TIDE()

tide.evaluate_range(dataset, bbox_results, mode=TIDE.BOX)

tide.summarize()
tide.plot()

[mrciolino]

Take a look at this project instead (for yolov8): I recently got it working

https://github.com/annsonic/identify_box_error