From: DLA-E: a deep learning accelerator for endoscopic images classification
No | Class / DNN Model Precision | Inception V3 | ResNet V2 | Inception_ResNet V2 | VGG19 | MobileNet V3 |
---|---|---|---|---|---|---|
1 | Pylorus | 0.91 | 0.92 | 0.93 | 0.91 | 0.95 |
2 | Cecum | 0.85 | 0.85 | 0.86 | 0.85 | 0.86 |
3 | Z-line | 0.87 | 0.84 | 0.93 | 0.96 | 0.95 |
4 | Esophagitis | 0.93 | 0.93 | 0.96 | 0.96 | 0.96 |
5 | Polyps | 0.97 | 0.99 | 0.96 | 0.88 | 0.94 |
6 | Ulcerative Colitis | 0.77 | 0.78 | 0.82 | 0.83 | 0.77 |
7 | Dyed and Lifted Polyps | 0.94 | 0.99 | 0.96 | 0.91 | 0.93 |
8 | Dyed Resection Margins | 0.97 | 0.96 | 0.98 | 0.96 | 0.97 |
Average Accuracy | 0.90 | 0.91 | 0.93 | 0.91 | 0.92 |