Table 3.

Setting of hyperparameters used during experiment.

Hyper parameters	Value
Model	Multichannel-Dense Net
No. of channel	04
Model initial learning rate	0.1
Image size	320 × 320 × 3
Batch size	04
Target labels	Ground Truth
Data augmentation	Not used
Loss function	Categorical cross-entropy
Optimization algorithm	Stochastic gradient decent
Validation parameter	Classification accuracy