Table 3.
Performance comparison of the YOLOv8-MU model and other models on the URPC2019 dataset.
| Model | AP (%) | mAP@0.5 (%) | Para (M) | FLOPs (G) | ||||
|---|---|---|---|---|---|---|---|---|
| Echinus | Starfish | Holothurian | Scallop | Waterweeds | ||||
| Boosting R-CNN [37] | 89.2 | 86.7 | 72.2 | 76.4 | 26.6 | 70.2 | 45.9 | 77.6 |
| Improved YOLOv5 [73] | 92.3 | 88.4 | 77.3 | 85.0 | 35.3 | 75.7 | 139.9 | 28.0 |
| YOLOv3 | 89.6 | 86.8 | 73.6 | 82.6 | 57.8 | 78.1 | 61.5 | 155.3 |
| YOLOv5s | 92.0 | 88.1 | 75.2 | 84.5 | 24.2 | 72.8 | 20.9 | 47.9 |
| YOLOv5m | 91.9 | 86.3 | 58.4 | 71.8 | 17.6 | 62.5 | 1.8 | 4.2 |
| YOLOv5n | 92.4 | 89.3 | 74.7 | 83.8 | 28.4 | 73.7 | 7.0 | 16.0 |
| YOLOv7 | 92.6 | 90.0 | 78.5 | 85.6 | 39.6 | 77.3 | 37.2 | 105.2 |
| YOLOv8s | 91.3 | 89.0 | 75.2 | 84.9 | 32.1 | 74.5 | 11.1 | 28.4 |
| YOLOv8m | 90.9 | 89.5 | 76.9 | 85.7 | 28.1 | 74.2 | 25.9 | 79.1 |
| YOLOv8l | 90.9 | 90.4 | 77.1 | 84.8 | 27.0 | 74.0 | 43.6 | 165.4 |
| YOLOv8n | 91.7 | 89.2 | 76.1 | 82.8 | 32.3 | 74.4 | 3.0 | 8.2 |
| YOLOv8-MU | 91.9 | 89.3 | 75.8 | 83.5 | 51.5 | 78.4 | 5.7 | 28.7 |