Table 8 Classification accuracy, precision, recall, and F1 score of the pre-trained model (fine-tuned) on the test set.

Models	Accuracy Mean (%)	Accuracy Std	Precision	Recall	F1-score	Training time (s)	Infer time (s)
ResNet-18	84.58	0.35	0.85	0.85	0.84	3.89	0.78
ResNet-34	81.41	0.54	0.81	0.81	0.81	3.51	0.14
ResNet-50	82.16	0.93	0.82	0.82	0.82	4.97	0.24
ResNet-101	77.51	1.03	0.78	0.78	0.77	8.78	0.41
ResNet-152	78.51	0.84	0.79	0.79	0.78	12.63	0.6
MobileNet-v1	85.62	2.16	0.87	0.86	0.85	2.2	0.14
MobileNet-v2	87.16	0.67	0.87	0.87	0.87	1.43	0.22
MobileNet-v3-small	72.28	1.64	0.73	0.72	0.72	1.33	0.22
MobileNet-v3-large	78.0	1.06	0.78	0.78	0.78	1.53	0.26
MobileNet-v4	87.74	1.23	0.88	0.88	0.87	3.0	0.53
EfficientNet-B0	89.52	4.07	0.9	0.9	0.89	11.2	0.65
EfficientNet-B1	86.5	4.44	0.87	0.86	0.86	11.2	0.65
EfficientNet-B2	85.38	5.79	0.85	0.85	0.85	11.2	0.65
EfficientNet-B3	85.63	3.89	0.86	0.86	0.86	11.2	0.65
EfficientNet-B4	81.95	5.31	0.82	0.82	0.82	11.2	0.65
EfficientNet-B5	86.55	4.14	0.87	0.87	0.86	11.2	0.65
EfficientNet-B6	82.34	4.75	0.82	0.82	0.82	11.2	0.65
EfficientNet-B7	84.91	3.46	0.85	0.85	0.85	11.2	0.65
EfficientNet-B8	88.34	2.96	0.88	0.88	0.88	11.2	0.65
EfficientNet-L2	88.74	1.91	0.89	0.89	0.89	11.2	0.65
RegNetY-200MF	83.82	2.31	0.84	0.84	0.84	5.41	0.32
RegNetY-400MF	81.41	0.38	0.81	0.81	0.81	5.41	0.32
RegNetY-600MF	86.1	1.63	0.86	0.86	0.86	5.41	0.32
RegNetY-800MF	89.12	1.87	0.89	0.89	0.89	5.41	0.32
RegNetY-1.6GF	85.43	1.04	0.85	0.85	0.85	5.41	0.32
RegNetY-3.2GF	87.59	0.75	0.88	0.88	0.87	5.41	0.32
RegNetY-4.0GF	83.42	0.93	0.83	0.83	0.83	5.41	0.32
RegNetY-6.4GF	84.98	1.35	0.85	0.85	0.85	5.41	0.32
RegNetY-8.0GF	87.94	1.33	0.88	0.88	0.88	5.41	0.32
RegNetY-12GF	90.35	1.0	0.9	0.9	0.9	5.41	0.32
RegNetY-16GF	88.66	1.77	0.89	0.89	0.89	5.41	0.32
RegNetY-32GF	91.34	1.33	0.91	0.91	0.89	5.41	0.32

Quick links

Search