TABLE 5.

Performance metric of the Drishti dataset using different CNN models.

CNN model	AUC	ACC	SEN	SP	PRE	FM	GM
ResNet50	0.9436	0.8898	0.9398	0.7955	0.8966	0.9176	0.8646
VGG19	0.9339	0.8740	0.9639	0.7045	0.8602	0.9091	0.8241
AlexNet	0.8771	0.8425	0.9880	0.5682	0.8119	0.8913	0.7492
Dns201	0.9480	0.8661	0.9277	0.7500	0.8750	0.9006	0.8341
IncRes	0.9102	0.8346	0.9277	0.6591	0.8370	0.8800	0.7819
Fusion	—	0.9055	0.7770	0.9259	0.8841	0.7770	0.8271

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