Table 8 Accuracy of proposed models with grayscale images and different brightness adjustments.

Model Name	0	+10%	+20%	+30%	− 10%	− 20%	-− 30%
AlexNet	0.721	0.723	0.733	0.730	0.709	0.700	0.688
EfficientNetV2	0.690	0.718	0.704	0.663	0.622	0.600	0.610
MobileNetV3	0.765	0.772	0.758	0.735	0.759	0.725	0.710
SqueezeNet	0.742	0.748	0.744	0.742	0.739	0.738	0.730
DenseNet201	0.745	0.756	0.741	0.727	0.735	0.710	0.695
ResNet101V2	0.723	0.742	0.741	0.742	0.722	0.690	0.654
ConvNeXt	0.733	0.741	0.737	0.730	0.715	0.690	0.670
DeepViT	0.725	0.724	0.732	0.725	0.715	0.702	0.680
LeViT	0.745	0.746	0.762	0.765	0.702	0.670	0.650
SwinTransformer	0.710	0.715	0.712	0.715	0.710	0.710	0.710
ViTbase	0.725	0.717	0.736	0.745	0.721	0.710	0.680
MaxViTsmall	0.772	0.767	0.767	0.766	0.781	0.772	0.740
Proposed Model	0.900	0.900	0.933	0.935	0.900	0.890	0.850

Quick links

Search