Table 5.

List of Hypermaters, loss and optimizer used for training classification and segmentation models.

Parameters	Segmentation model	Classification model
Batch size	32	64
Learning Rate (Initial)	0.0001	0.0001
Epochs	45	50
Shuffle Each Iteration	Yes	Yes
Stopping criteria	5	10
Loss function	Negative Log Likelihood Loss/MSE	CELoss
Optimizer	SGDM	SGDM