Table 6.

Comparison of $F_1$ scores of CNN models at various image resolutions for whole wheat bread.

CNN Model	Image Resolution (pixels)
	224 $\times$ 224	448 $\times$ 448	672 $\times$ 672	896 $\times$ 896
DenseNet121	0.9653	0.9856	0.9906	0.9947
EfficientNetV2B0	0.9750	0.9903	0.9909	0.9954
EfficientNetV2B1	0.9826	0.9879	0.9941	0.9972
EfficientNetV2B2	0.9887	0.9844	0.9903	0.9910
EfficientNetV2B3	0.9718	0.9846	0.9868	0.9972
EfficientNetV2L	0.9379	0.9873	0.9923	0.9906
EfficientNetV2M	0.9341	0.9748	0.9880	0.9941
EfficientNetV2S	0.9726	0.9887	0.9924	0.9960
InceptionResNetV2	0.8996	0.9399	0.9619	0.9649
InceptionV3	0.9091	0.9600	0.9699	0.9770
MobileNet	0.8976	0.9768	0.9776	0.9893
MobileNetV2	0.9243	0.9662	0.9852	0.9861
MobileNetV3Large	0.9637	0.9838	0.9872	0.9972
MobileNetV3Small	0.9659	0.9866	0.9932	0.9972
NASNetMobile	0.8687	0.9126	0.9511	0.9441
RegNetX016	0.9808	0.9896	0.9912	0.9853
RegNetX032	0.9856	0.9895	0.9887	0.9892
RegNetY016	0.9842	0.9931	0.9960	0.9947
RegNetY032	0.9858	0.9941	0.9972	0.9972
ResNet50	0.9825	0.9775	0.9944	0.9944
ResNet50V2	0.9054	0.9540	0.9657	0.9816
Xception	0.8881	0.9608	0.9787	0.9870