Table 2. Inference speed of the four light weight and two heavy weight models along with other parameters which affects the inference speed.

Name of model	Inference time (ms)	Number of layers	Number of params	Model size (MB)
UNet	42	32	7,759,521	30
SegNet	84	64	31,819,649	122
VGG UNet	65	85	25,882,433	99
HR Net	140	1043	28,607,456	112
Inception ResNet V2	123	572	55,873,736	215
ResNet 101	115	101	44,675,560	171