Autofocus for semseg?
I have not seen people for whom DeepLab worked...and in my tests dilated convolutions were the same...though some claim they help with high-res images with small objects...
(0) Autofocus layer, a novel module that enhances the multi-scale processing of CNNs by learning to select the ‘appropriate’ scale for identifying different objects in an image
(1) Layer description
(2) Implementation github.com/yaq007/Autofocus-Laye
I believe this will work best for 3D images