Table 2.

The experimental results of cassava agricultural diseases and pests.

Methods		Accuracy (%)	Precision (%)	Recall (%)	F1
Coarse-grained	VGG19 [32]	88.0	79.0	77.8	0.78
	ResNet50 [33]	89.7	77.0	74.5	0.76
	Inception [34]	88.5	79.3	78.2	0.78
	DenseNet [48]	89.1	79.6	78.4	0.79
	CSPNet50 [49]	92.3	80.1	80.1	0.81
	Senet [50]	94.5	84.3	80.2	0.82
Fine-grained	LIR [51]	98.5	88.3	86.6	0.88
	WSAP [52]	98.4	88.4	87.2	0.88
	FBSD [24]	98.6	87.2	86.5	0.87
	GHA-Net	97.4	89.2	87.9	0.90