Table 8 A relative performance comparison of models trained on the merged dataset against the GWHD baseline.
From: Transfer learning models for wheat ear detection on multi-source dataset
Model | \(\Delta\)mAP@50 | \(\Delta\) Training duration |
|---|---|---|
F-RCNN R 50 | − 3.89 | +08min 38s |
F-RCNN R 101 | − 12.71 | +11min 20s |
F-RCNN X 101 32x8d | − 0.51 | +30min 04s |
RetinaNet R 101 | +3.00 | +09min 58s |
YOLOv8 | +2.09 | +21min 53s |
RT-DETR | +3.53 | +26min 40s |
