SERR-U-Net: Squeeze-and-Excitation Residual and Recurrent Block-Based U-Net for Automatic Vessel Segmentation in Retinal Image

Table 5

Comparative results between our proposed method and expert 2^nd.

Database	Method	ACC	AUC	SP	SE
STARE	Ours	0.9796	0.9859	0.9926	0.8220
	Expert 2^nd	0.9347	-	0.9382	0.8955
DRIVE	Ours	0.9552	0.9784	0.9813	0.7792
	Expert 2^nd	0.9464	-	0.9717	0.7796