Table 1. RNGU-NET model architecture.

Layer	Channel size	Operators	Layer	Input size (x)	Operators
L-Stage1	256*256*(32)	Down conv, ReLU, ResNet	R-Stage1	256*128*(128)	Up conv, ReLU,GAV
L-Stage2	256*256*(64)	Down conv, ReLU,ResNet	R-Stage2	128*64*(64)	Up conv, ReLU,GAV
L-Stage3	256*256*(128)	Down conv, ReLU,ResNet	R-Stage3	64*32*(32)	Up conv, ReLU,GAV
L-Stage4	256*256*(256)	Down conv, ReLU,ResNet	R-Stage4	32*32*(32)	Upconv, ReLU,GAV
L-Stage5	256*256*(256/2)	NLB	R-Stage5	1*1*(32)	Conv-sigmoid (output)