Table 1.

Model hyperparameter settings

Model	Learning Rate	Optimizer	Learning Rate Scheduler	Batch Size	Epochs	Patience
EfficientNet	1e−4	Adam	StepLR	60	80	8
ResNet50	1e−4	Adam	StepLR	60	80	8
VGG19	1e−3	SGD	StepLR	60	80	8
DenseNet121	1e−3	Adam	StepLR	60	80	8
ViT	1e−3	Adam	StepLR	32	80	8
Swin Transformer	1e−5	AdamW	StepLR	64	80	8
CvT	1e−5	AdamW	StepLR	64	80	8