Table 12. Numerical results obtained from models using the APTOS 2019 Blindness Detection test dataset.

Model	Metric	AF						AFCS-CNN
		ReLU	ELU	SELU	Mish	SiLU	GELU
VGG16	Accuracy	0.7709	0.7381	0.7599	0.7945	0.7945	0.7927	0.8254
	Loss	1.0019	0.7306	0.6178	0.6689	0.5773	0.9456	0.4848
	Precision	0.7589	0.7468	0.6838	0.7846	0.8030	0.7921	0.8196
	Recall	0.7693	0.7390	0.7597	0.7970	0.7923	0.7935	0.8234
	F1-score	0.7601	0.7308	0.7170	0.7838	0.7975	0.7901	0.8130
VGG19	Accuracy	0.7981	0.7727	0.7418	0.8090	0.7545	0.8000	0.8181
	Loss	0.5737	0.6036	0.6501	0.6260	0.6962	0.5305	0.5149
	Precision	0.7925	0.7231	0.7238	0.7961	0.7154	0.7902	0.8132
	Recall	0.7971	0.7744	0.7413	0.8078	0.7532	0.8008	0.8173
	F1-score	0.7943	0.7368	0.7209	0.8013	0.7081	0.7939	0.8092
DenseNet121	Accuracy	0.8072	0.7654	0.7690	0.7818	0.7327	0.7854	0.8327
	Loss	0.7971	0.6421	0.5645	0.5828	1.0582	0.6388	0.5704
	Precision	0.8080	0.8246	0.7714	0.8173	0.7720	0.7886	0.8335
	Recall	0.8066	0.7634	0.7687	0.7816	0.7325	0.7872	0.8329
	F1-score	0.8049	0.7611	0.7607	0.7877	0.7251	0.7817	0.8135
DenseNet169	Accuracy	0.7836	0.6218	0.5836	0.7290	0.7618	0.8163	0.8254
	Loss	1.0642	1.0636	1.2094	0.9553	0.6982	0.7083	0.6344
	Precision	0.7825	0.6644	0.6388	0.7760	0.7766	0.8182	0.8258
	Recall	0.7813	0.6203	0.5853	0.7286	0.7649	0.8196	0.8233
	F1-score	0.7777	0.6185	0.5693	0.7173	0.7580	0.8027	0.8071
EfficientNetV2B0	Accuracy	0.6363	0.7490	0.5690	0.7454	0.7163	0.7127	0.8054
	Loss	1.0799	0.8925	1.1607	1.0322	1.5423	1.0767	0.8327
	Precision	0.6663	0.7050	0.6519	0.7261	0.7260	0.6596	0.7985
	Recall	0.6383	0.7486	0.5675	0.7461	0.7159	0.7113	0.8076
	F1-score	0.6448	0.7124	0.5953	0.7277	0.6797	0.6634	0.8013
EfficientNetV2B1	Accuracy	0.6836	0.6399	0.6454	0.7927	0.7309	0.7363	0.8254
	Loss	1.0460	1.2034	1.1211	1.1672	1.0322	1.0834	0.8308
	Precision	0.6712	0.5901	0.6211	0.7718	0.7080	0.7365	0.8269
	Recall	0.6864	0.6378	0.6473	0.7909	0.7284	0.7374	0.8251
	F1-score	0.6392	0.6097	0.6188	0.7657	0.7061	0.7077	0.8194