Additional tricks / post processing steps?

[mrgloom]

I wonder regarding geospatial semantic segmentation is there any tricks/post processing steps to make objects like house have sharp box like shape(i.e. naive solution give smooth shapes without sharp corners).