More masks than bboxes?

Food Recognition Challenge
#1

Some annotations appear to have more segmentation masks than bounding boxes. For instance the instance 66 of the train dataset has 2 masks, but only 1 bbox.

i = 66
xx = annotations["annotations"][i]['segmentation']
yy = annotations["annotations"][i]['bbox']
print(xx)
print(len(xx))
print(yy)
print(len(yy))


[[163.90875, 115.78999999999999, .... 49999999996]]
 2
[118.79625, 324.30999999999995, 270.17375, 376.44]
4
#2

Hi, thanks for pointing this out! We are looking into this issue and giving updates asap :slight_smile: