“…The success of training a deep convolutional neural network (CNN) [ 55 ] on large scale object recognition dataset [ 136 ] has had a huge impact on the entire research community. Researchers from diverse fields such as natural language processing [ 3 , 137 ], computer networks [ 138 , 139 , 140 ], stock market analysis [ 141 , 142 ], document analysis and recognition [ 143 , 144 ], and of course computer vision [ 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 ] have effectively leveraged deep-learning to devise models that achieve appreciable performance compared to heuristic, and classical machine learning (ML) techniques. Following other related fields in computer vision, existing pre-trained CNNs [ 56 , 145 ] for image classification task on ImageNet dataset have been re-purposed to effectively address various challenges present in the SOD task.…”