plot_mask() method failing when expand_mask() returns mask with just one object.

[sunny2309]

Hello MXNet Team,

I was recently trying Image segmentation (instance segmentation) using Pre-trained models available through BluonCV.

I used a simple image with just a panda (https://upload.wikimedia.org/wikipedia/commons/thumb/3/3c/Giant_Panda_2004-03-2.jpg/1200px-Giant_Panda_2004-03-2.jpg) in it.

I used "mask_rcnn_resnet101_v1d_coco" model for Instance segmentation. It works fine and detects one object which is pandas.

After that, I used expand_mask() method from viz utils to expand masks. It works fine.

But when I call plot_mask() method with an output of expand_mask() which has only one object (panda) in our case, it fails with an array indexing error.

IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed

The method works fine when the image has more than one object detected. It fails only when it has just one object detected.

I simply repeated the same mask twice by using the mxnet stack() method and then the repeated mask worked fine with plot_mask() method.

panda_masks_mod = mxnet.nd.stack(mxnet.nd.array(panda_masks_mod), mxnet.nd.array(panda_masks_mod), axis=1)

Regards, Sunny

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.