Table 5 Other researchers’ results.
From: A deep learning-based approach for axle counter in free-flow tolling systems
Author | Application | Model | Precision | Recall | mAP @0.50 | mAP @0.50:0.95 |
|---|---|---|---|---|---|---|
Stefenon et al.55 | Insulator | YOLOv5u | 0.981 | 0.975 | 0.983 | 0.905 |
Souza et al.56 | Insulator | YOLOv5x | 0.983 | 0.990 | 0.993 | 0.957 |
Han et al.58 | Vehicle location | O-YOLO-v2 | 0.940 | 0.940 | 0.940 | – |
Padilla Carrasco et al.57 | Vehicle location | T-YOLO | 0.963 | 0.996 | 0.998 | 0.997 |
Li et al.53 | Vehicle wheels | YOLOv3-SPP | 0.718 | – | 0.955 | – |
Miles et al.52 | Vehicle wheels | YOLOv3 | 0.927 | – | 0.927 | – |
Our | Vehicle wheels | YOLOv5m | 0.994 | 0.982 | 0.994 | 0.800 |
