Table 10 Comparison of performance evaluation indicators of different object detection models on the RDD2022 test set.
From: Object detection model design for tiny road surface damage
Models | AP50(%) | P(%) | R(%) | mAP50(%) | |||
|---|---|---|---|---|---|---|---|
Longitudinal cracks | Transverse cracks | Alligator cracks | Potholes | ||||
YOLOv8n | 55.3 | 55.9 | 66.8 | 42.3 | 62.6 | 50.8 | 55.1 |
YOLOv8s | 57.2 | 59.0 | 68.5 | 46.8 | 62.7 | 54.0 | 57.9 |
YOLOv8m | 59.0 | 60.0 | 68.6 | 47.8 | 64.7 | 54.3 | 58.9 |
YOLOv8l | 59.5 | 61.0 | 68.4 | 48.5 | 64.9 | 54.6 | 59.4 |
YOLOv8x | 59.8 | 61.2 | 68.5 | 49.5 | 64.2 | 56.2 | 59.7 |
YOLOv6n | 54.1 | 55.3 | 65.5 | 38.5 | 60.8 | 49.8 | 53.3 |
YOLOv6s | 57.4 | 58.1 | 67.3 | 45.2 | 64.6 | 51.6 | 57.0 |
YOLOv6m | 55.5 | 56.9 | 65.7 | 45.2 | 63.1 | 51.9 | 55.8 |
YOLOv6l | 54.6 | 55.6 | 63.6 | 39.2 | 61.1 | 49.4 | 53.2 |
YOLOv5n | 54.3 | 53.9 | 64.3 | 40.4 | 60.7 | 49.9 | 53.2 |
YOLOv5s | 57.0 | 57.0 | 67.5 | 43.7 | 63.0 | 51.2 | 56.3 |
YOLOv5m | 59.0 | 53.7 | 65.3 | 48.5 | 61.7 | 54.6 | 56.6 |
YOLOv5l | 59.6 | 56.5 | 65.1 | 50.3 | 63.9 | 54.7 | 57.9 |
YOLOv5x | 60.1 | 59.9 | 69.9 | 50.4 | 64.4 | 55.7 | 60.1 |
YOLOv9 (gelan) | 62.5 | 60.8 | 69.1 | 49.7 | 65.5 | 55.6 | 60.5 |
YOLOv9 (gelan-c) | 62.0 | 60.9 | 68.5 | 49.8 | 65.4 | 55.2 | 60.3 |
Ours | 60.3 | 59.2 | 71.5 | 53.6 | 66.7 | 56.0 | 61.2 |
