Table 7.

Performance of competing models for ES and non-ESED frames.

Competing approaches	(Uncropped images)				(Cropped images)
	AUC	Loss	Precision	Recall	AUC	Loss	Precision	Recall
DeepPhase	0.73	0.2	0.63	0.7	0.8	0.15	0.78	0.75
CNN proposed by Siddiqi27	0.45	0.52	0.52	0.5	0.68	0.27	0.6	0.58
VGG16	0.65	0.3	0.56	0.59	0.68	0.27	0.6	0.62
MobileNet	0.67	0.28	0.64	0.65	0.7	0.26	0.67	0.68
EfficientNet	0.64	0.27	0.55	0.6	0.68	0.25	0.62	0.68
ResNet	0.63	0.3	0.5	0.55	0.71	0.28	0.67	0.69
Lane et al.20	0.68	0.22	0.6	0.62	0.71	0.16	0.72	0.7
Wegner et al.21	0.63	0.22	0.61	0.63	0.76	0.22	0.7	0.72