Myriad efforts have been made over the last 10 years in algorithmic improvements and dataset creation for semantic segmentation tasks. Of late, there have been rapid gains in this field, a subset of visual scene understanding, due mainly to contributions by deep learning methodologies. But deep ...