Table 4 Evaluators of the goodness of fit for RCNNS with different architectures in classifying sweet potato roots according to shape, damage caused by insects, and skin color. UFMG (2022).

Var	Architecture	Precision	Recall	F1	Accuracy	Specificity
Shape	VGG-16	0.797	0.803	0.800	0.776	0.743
	Inception-v3	0.775	0.843	0.807	0.776	0.693
	ResNet-50	0.767	0.817	0.791	0.760	0.689
	InceptionResNetV2	0.969	0.978	0.974	0.971	0.961
	EfficientNetB3	0.804	0.880	0.841	0.814	0.732
Damage by insects	VGG-16	0.245	0.678	0.360	0.702	0.705
	Inception-v3	0.529	0.776	0.629	0.887	0.902
	ResNet-50	0.225	0.510	0.313	0.723	0.752
	InceptionResNetV2	0.787	0.979	0.872	0.965	0.963
	EfficientNetB3	0.249	0.608	0.354	0.725	0.742
Skin color	VGG-16	0.869	0.829	0.848	0.857	0.883
	Inception-v3	0.712	0.813	0.759	0.750	0.692
	ResNet-50	0.831	0.859	0.845	0.847	0.836
	InceptionResNetV2	0.996	0.964	0.980	0.981	0.997
	EfficientNetB3	0.903	0.902	0.903	0.906	0.910

Quick links

Search