Table 5 Summary of results for gastric polyps detection
From: Gastric polyp detection module based on improved attentional feature fusion
Method | F1 | P (%) | R (%) |
|---|---|---|---|
CSP-Darknet53 | 0.76 | 84.00 | 70.00 |
Faster R-CNN | 0.84 | 88.10 | 80.90 |
SSD | 0.79 | 77.00 | 80.90 |
YOLOv3 | 0.81 | 87.80 | 75.70 |
YOLOv4 | 0.87 | 93.80 | 80.78 |
YOLOv5s | 0.88 | 85.60 | 91.90 |
CSP-Darknet53+our attentional feature fusion module | 0.89 | 95.92 | 83.63 |