PRE_NMS_LIMIT

[yuriyOn]

I got an error while running first evaluate.ipynb and then train.ipynb that PRE_NMS_LIMIT is not defined in the model.py. I have found in the original Mask R-CNN implementation that it should be set to 6000. You should add the parameter to avoid the error to both files.

[michaelisc]

It seems you are using a more recent version of the Matterport Mask R-CNN than included in the repo. The PRE_NMS_LIMIT was fixed to 6000 back then. Can you test if any other things fail with the newer version of Mask R-CNN? If this is the only thing we have to change I am happy to add this.

[yuriyOn]

Another problem is listed in issue #1 and it is solved by replacing the Mask_RCNN directory with the latest version.

[michaelisc]

The problem with PRE_NMS_LIMIT should be fixed with commit 2a677a9. Now the settings from the original Config class from Mask_RCNN/mrcnn/config.py are inherited.

[michaelisc]

We should still update lib/Mask_RCNN. See issue #5