Test models trained on my own data, many objects only have detection box without mask

matterport / Mask_RCNN

Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow

Other

24.7k stars 11.71k forks source link

Test models trained on my own data, many objects only have detection box without mask #1696

Open LiXm1002 opened 5 years ago

LiXm1002 commented 5 years ago

Hello everyone， I have problem when testing models trained on my own data, but many objects only have detection box without mask. Is my training time not enough, or the parameters are incorrect?

JOSHUAWENJZ commented 5 years ago

I just have the same problem, please someone can solve it.

Hunain-Gillani commented 5 years ago

What sort of annotation are you using? Make sure you're using polygon or bitmap annotation which Masks the object