Table 5.

Comparison with well-known CNN-based classifiers on patch-based MIM datasets.

	Patch-Based MIM Dataset
CNN-Based Classifier	Medium RAFT-MIM${}^{{\mathit{Q}}_{12}}$		Small RAFT-MIM${}^{{\mathit{Q}}_{12}}$		Medium RAFT-MIM${}^{{\mathit{Q}}_{25}}$		Small RAFT-MIM${}^{{\mathit{Q}}_{25}}$
	Accuracy%	F1 Score%	Accuracy%	F1 Score%	Accuracy%	F1 Score%	Accuracy%	F1 Score%
MobileNet	87	87	79	78	85	85	77	74
EfficientNet-B0	88	88	81	80	87	87	78	78
InceptionV3	85	85	76	75	80	80	76	74
ResNet50	80	80	70	70	74	73	71	69

Bold: best results in each dataset. Gray highlight: Best results among all datasets.