Table 2.

Classification performance of different methods on Cataract101

Methods	Accuracy	Precision	Recall	Jaccard
ResNet50 [33]	0.864	0.828	0.823	0.710
Qi et al. [34]	0.881	-	-	-
OHFM [23]	0.920	0.892	0.903	0.816
SV-RCNet [22]	0.934	0.913	0.922	0.848
TeCNO [20]	0.936	0.917	0.915	0.847
STANet [21]	0.953	0.934	0.935	0.879
GL-MSTCN	0.965	0.949	0.952	0.908