Table 5 yolov8 comparison of adding CPAC to each layer.
From: Optimized YOLOv8 framework for intelligent rockfall detection on mountain roads
Model | AP@0.5 | AP@0.75 | parameters | GFLOPs |
|---|---|---|---|---|
Yolov8 + cpc Shallow | 0.879 | 0.502 | 3,008,023 | 8.5 |
Yolov8 + cpc Medium | 0.882 | 0.502 | 3,019,171 | 8.2 |
Yolov8 + cpc Deep | 0.881 | 0.508 | 3,132,883 | 8.2 |
