Table 2.

Comparison of Counting Performance for Different Models.

Model	MAE	RMSE	MAPE	R2
YOLOv8	6.16	8.89	10.05%	0.9035
CenterNet	12.67	17.77	18.75%	0.6926
Faster R-CNN	16.56	19.22	34.95%	0.7211
TasselLFANet	5.92	8.83	9.82%	0.9002
EfficientDet	22.74	38.41	29.17%	0.2777
FCOS	11.88	22.00	16.43%	0.6537
YOLO POD	4.18	10.04	6.48%	0.9666
PodNet	4.52	7.65	6.24%	0.9500

The best performance is in boldface.