Table 5 Experimental results of YOLO based detection method on Seaships 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 | 93.6 | 93.5 | 95.2 | 78 | 1.9 |
YOLOv6 | 92.5 | 91.8 | 96.6 | 69.6 | 4.3 |
YOLOv7 | 95.4 | 93.7 | 93.7 | 79.8 | 3.2 |
YOLOv9 | 95.7 | 94 | 96.4 | 76.5 | 3.3 |
YOLOv10 | 92.4 | 94.9 | 94.7 | 80.8 | 2.7 |
YOLOv11 | 96.1 | 98.7 | 95.3 | 76.5 | 3.3 |
Faster-RCNN11 | 95.5 | 96.8 | 94.8 | 77.6 | 100 |
YOLO-MS35 | 98.2 | 93.1 | 94.1 | 78.4 | 2.5 |
YOLO-NL36 | 90.3 | 89.4 | 91.6 | 64.2 | 4.5 |
YOLO-PL37 | 91.5 | 90.2 | 92.3 | 64.6 | 3.5 |
Gold-YOLO38 | 93.4 | 96.8 | 94.6 | 67.9 | 4.1 |
FFCA-YOLO39 | 94.7 | 91.4 | 95.8 | 67.4 | 4.6 |
SOD-YOLO40 | 95.3 | 95.8 | 96.3 | 71.6 | 4.5 |
YOLO-HPSD41 | - | - | 98.8 | - | - |
GEW-YOLO(ours) | 97.6 | 98.1 | 99.1 | 86.3 | 1.2 |
