Table 2.

Formulae for model evaluation metrics.

Abbreviation	Full Name or Calculation Formulas	Serial Number
TP	Positive samples predicted as positive by the model (correctly predicted positive samples)	(1)
TN	Negative samples predicted as negative by the model (correctly predicted negative samples)	(2)
FP	Negative samples predicted as positive by the model (positive samples with incorrect predictions)	(3)
FN	Positive samples predicted as negative by the model (negatively classified samples with incorrect predictions)	(4)
Iou	$\text{IoU}=\frac{\text{TP}}{\text{TP}+\text{FP}+\text{FN}}$	(5)
Accuracy	$\text{Acc}=\frac{\text{TP}+\text{TN}}{\text{TP}+\text{TN}+\text{FP}+\text{FN}}$	(6)
F1-score	$\text{F}1=2*\frac{\text{P}*\text{R}}{\text{P}+\text{R}}$	(7)
Precision	$\text{P}=\frac{\text{TP}}{\text{TP}+\text{FP}}$	(8)
Recall	$\text{R}=\frac{\text{TP}}{\text{TP}+\text{FN}}$	(9)