Research on Classification of Fine-Grained Rock Images Based on Deep Learning

Table 5

Comparison of classification accuracy of different algorithms under the test set.

Method	Model	Accuracy (%)
Original	SE-ResNet-50	40.625
	ViT	50.0
	EfficentNet-B0	31.25
Image cutting	SE-ResNet-50	28.125
	ViT	43.75
	EfficentNet-B0	34.375
SBV	SE-ResNet-50	71.875
	ViT	56.25
	EfficentNet-B0	75.0
Image cutting + SBV	SE-ResNet-50	75.0
	ViT	68.75
	EfficentNet-B0	75.0