Object Detection on Weed Dataset 2021-2023
65.2PrecisionDINO*-YOLO26*
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DINO*-YOLO26*Backbone=Fine-tuned DINO-ViT, Training loss=DFL loss2026.02 | 65.2 | 54.4 | 56.5 | 33.1 | |
| DINO-YOLO26*Backbone=DINOv3-pretrained ViT, Training loss=DFL loss2026.02 | 63.9 | 49.9 | 51.8 | 31 | |
| YOLO26*Training loss=DFL loss2026.02 | 58.4 | 44.8 | 46.3 | 28.2 | |
| YOLO26Training loss=Standard2026.02 | 55.1 | 42.9 | 42.5 | 25.7 |