Table 4.

Summary performance index of transfer learning networks.

Transfer Learning Type	CNN	Batch Size	Accuracy	Precision	Recall	F1-Score	Time (Min)
Fine-Tuning	Resnet18	16	0.8517	0.98	0.60	0.75	12.7
		32	0.8467	0.96	0.59	0.73	12.7
	Resnet50	16	0.8628	0.99	0.63	0.77	17.2
		32	0.8576	0.99	0.61	0.75	17.2
Fixed Feature Extraction	Resnet18	16	0.7849	0.84	0.48	0.61	12.5
		32	0.7849	0.86	0.44	0.58	12.7
	Resnet50	16	0.7878	0.88	0.45	0.59	12.7
		32	0.7878	0.90	0.43	0.58	12.7