Table 2 Comparison of evaluation indexes of VGG19, Inception-v3, Resnet50, Mobilenet, Xception and ACCV
From: ACCV: automatic classification algorithm of cataract video based on deep learning
Method | Acc | Sensitivity | Specificity | Pre | F1 |
|---|---|---|---|---|---|
ACCV | 0.9400 | 0.9200 | 0.9600 | 0.9580 | 0.9388 |
Mobilenet | 0.8800 | 0.8200 | 0.9400 | 0.9318 | 0.8723 |
VGG-19 | 0.8700 | 0.7600 | 0.9800 | 0.9744 | 0.8539 |
Inception-v3 | 0.8100 | 0.6600 | 0.9600 | 0.9429 | 0.7765 |
ResNet-50 | 0.8600 | 0.8000 | 0.9200 | 0.9091 | 0.8511 |
Xception | 0.8600 | 0.8200 | 0.9000 | 0.8943 | 0.8542 |