Table 4 Details of the comparison with... | Emerald Publishing

Table 4

Details of the comparison with the existing works that have used the same dataset

Evaluation method	Reference	Architecture	Accuracy (%)	Recall (%)	Precision (%)	F1-score (%)
80/20% test split	[2]	CNN	94.55	96.5	96.0	96.0
	[37]	EfficientNetB1 + ResNet50	95.98	95.98	96.0	95.98
	Ours	ResNet-CBAM	99.43	99.0	98.7	99.0
Training and testing data	[38]	CNN	95.65	95.65	95.67	95.65
Training and testing data	Ours	ResNet-CBAM	99.15	98.16	98.42	98.29
60/20/20	[39]	VGG19	97.00	96.0	97.0	97.0
60/20/20	Ours	ResNet-CBAM	98.53	96.76	97.38	97.06
5-fold CV	[14]	CNN	98.40	–	96.75	96.75
5-fold CV	Ours	ResNet-CBAM	99.35	98.55	98.90	98.70

Evaluation method	Reference	Architecture	Accuracy (%)	Recall (%)	Precision (%)	F1-score (%)
80/20% test split	[2]	CNN	94.55	96.5	96.0	96.0
	[37]	EfficientNetB1 + ResNet50	95.98	95.98	96.0	95.98
	Ours	ResNet-CBAM	99.43	99.0	98.7	99.0
Training and testing data	[38]	CNN	95.65	95.65	95.67	95.65
Training and testing data	Ours	ResNet-CBAM	99.15	98.16	98.42	98.29
60/20/20	[39]	VGG19	97.00	96.0	97.0	97.0
60/20/20	Ours	ResNet-CBAM	98.53	96.76	97.38	97.06
5-fold CV	[14]	CNN	98.40	–	96.75	96.75
5-fold CV	Ours	ResNet-CBAM	99.35	98.55	98.90	98.70

Source(s): Table created by the authors