DM_CorrMatch: a semi-supervised semantic segmentation framework for rapeseed flower coverage estimation using UAV imagery

Table 2 Comparison of ablation experiment accuracy

Base Network	Mamba	AUL	Diffusion	Precision	Recall	IoU	F1 score	Accuracy	Flops (G)	Parameters (M)
\(\checkmark\)				0.910	0.913	0.865	0.911	0.922	66.86	64.18
\(\checkmark\)	\(\checkmark\)			0.923	0.924	0.874	0.924	0.923	73.87	66.74
\(\checkmark\)		\(\checkmark\)		0.923	0.922	0.872	0.922	0.923	67.96	65.12
\(\checkmark\)	\(\checkmark\)	\(\checkmark\)		0.945	0.948	0.893	0.946	0.945	74.97	67.68
\(\checkmark\)			\(\checkmark\)	0.931	0.925	0.875	0.931	0.930	85.33	73.28
\(\checkmark\)	\(\checkmark\)	\(\checkmark\)	\(\checkmark\)	0.942	0.940	0.886	0.941	0.940	84.54	76.78

Bold values highlight the best results obtained in the ablation study in terms of accuracy and model complexity

ISSN: 1746-4811