Table 4 Experimental results of YOLO based detection method on Dockship dataset.
From: Multi-ship detection and classification with feature enhancement and lightweight fusion
Model | Precision(%) | Recall(%) | mAP@0.5(%) | mAP@0.5:0.95(%) | Params(M) |
|---|---|---|---|---|---|
YOLOv5 | 78.2 | 73.1 | 78.4 | 50.2 | 1.9 |
YOLOv6 | 78.7 | 74.7 | 79.5 | 51.4 | 4.3 |
YOLOv7 | 80.2 | 74.2 | 79.2 | 52.1 | 3.2 |
YOLOv9 | 80.6 | 69.5 | 76.4 | 49.2 | 3.3 |
YOLOv10 | 80.6 | 75.6 | 80.8 | 52.0 | 2.7 |
YOLOv11 | 79.3 | 76.7 | 81.2 | 53.5 | 3.3 |
Faster-RCNN11 | 80.1 | 73.4 | 79.8 | 51.6 | 100 |
YOLO-MS35 | 81.6 | 73.0 | 81.0 | 51.9 | 2.5 |
YOLO-NL36 | 74.7 | 71.7 | 72.5 | 46.7 | 4.5 |
YOLO-PL37 | 73.2 | 74.1 | 72.6 | 45.4 | 3.5 |
Gold-YOLO38 | 75.8 | 70.2 | 74.5 | 46.8 | 4.1 |
FFCA-YOLO39 | 73.6 | 71.2 | 73.9 | 47.4 | 4.6 |
SOD-YOLO40 | 75.7 | 74.4 | 74.5 | 46.9 | 4.5 |
GEW-YOLO(ours) | 79.4 | 76.9 | 82.1 | 61.1 | 1.2 |
