Table 4.

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

Competing approaches	AUC	Loss	Precision	Recall
DeepPhase	0.85	0.15	0.78	0.72
CNN proposed by Siddiqi27	0.7	0.31	0.62	0.65
VGG16	0.68	0.27	0.6	0.62
MobileNet	0.74	0.26	0.65	0.68
EfficientNet	0.76	0.25	0.7	0.72
ResNet	0.71	0.28	0.7	0.69
Lane et al.20	0.8	0.17	0.7	0.72
Wegner et al.21	0.81	0.27	0.71	0.73