Spritaro

I tried to use this PR for a single-class instance segmentation task, but the trained model produced more false positives. I looked up mrcnn_class_loss_graph and noticed that ROIs are removed...

@konstantin-frolov It's from files changed in this pull request. You can also find it here https://github.com/keineahnung2345/Mask_RCNN/blob/168ca5cdb7f5656722ec6ed2dd67451cde994421/mrcnn/model.py#L1118

I had a similar problem. I was trying to copy weights from model1 to model2 and convert the model2 to ONNX, then it raised the error ```must be str, not...

It seems ```sigmoid + norm``` is used by default. Adding ```MODEL.SPARSE_INST.DECODER.NAME GroupIAMSoftDecoder``` to the command line solved the problem for me.