*** No instances to display *** Error while doing inference

[atinesh]

Hello, I have trained the model on custom dataset for 5 epochs, but while doing inference I am getting below error

I tried the following modifications in the code but error is not resolved

1. Lowered DETECTION_MIN_CONFIDENCE in mrcnn/config.py Issue: 2132

2. Fixed Division by Zero error in mrcnn/utils.py with these modifications from __future__ import division Issue: 515, shift = np.array([0, 0, 1., 1.]) Issue: 912

[raditharch]

hello, I got the same problem and I use same solution as you, but I added scipy versions less than 1.3.0 and it's success.

[raditharch]

here, the reference : matterport/Mask_RCNN#2132

[zacialen]

Hi, were you able to fix the problem? I tried the solutions above, but I am still not getting any. Any suggestions?