TABLE 6.

Performance metric of the HVD + Drishti dataset using different CNN models.

CNN model	AUC	ACC	SEN	SP	PRE	FM	GM
ResNet50	0.9671	0.8288	0.8636	0.8118	0.6924	0.7686	0.8373
VGG19	0.9653	0.8306	0.8855	0.8037	0.6888	0.7749	0.8346
AlexNet	0.9479	0.7917	0.9455	0.7163	0.6205	0.7493	0.8230
Dns201	0.9671	0.8145	0.8400	0.8020	0.6754	0.7488	0.8208
IncRes	0.9318	0.7522	0.7600	0.7484	0.5971	0.6688	0.7542
Fusion	—	0.8518	0.7992	0.9252	0.7969	0.7992	0.7980

Bold values are showing the model results after applying classifier fusion operation.