Table 2.

YOLO models hyperparameter setting for training and validation.

Model	Input size	Batch size	Decay	Momentum	Learning rate
YOLOv7-W6	640×640	12	0.005	0.937	0.1
YOLOv7-D6	640×640	12	0.005	0.937	0.1
YOLOv7x	640×640	12	0.005	0.937	0.1
YOLOv8s	640×640	20	0.001	0.937	0.01
YOLOv8m	640×640	20	0.001	0.937	0.01
YOLOv8n	640×640	20	0.001	0.937	0.01
YOLOv8l	640×640	20	0.001	0.937	0.01
YOLOv8x	640×640	20	0.001	0.937	0.01