Table 3.

Hyper-parameter training parameters for the proposed model.

Parameter	Characteristics	Value
Batch size	If we tuning the batch size value, the grouping of each trained value increases the efficiency of the model	16
Epochs	Easy to learn underlying different patterns	50
Learning rate	Decreases the computational efficiency of layers	0.001
Minimum Threshold value for IoU	Helps to detect the segmented regions with no overlapping	0.5
Optimizer	Minimize the loss in segmentation regions	Adam
Weight decay rate	Increases the learning rate	0.005