Table 1 Quantitative comparison of spore segmentation performance for different model architectures.

From: Attention-driven UNet enhancement for accurate segmentation of bacterial spore outgrowth in microscopy images

Method

Precision

Sensitivity

Specificity

F1-score

Accuracy

IOU

Cellpose

0.78

0.64

0.78

0.68

0.70

0.54

Omnipose

0.79

0.67

0.88

0.73

0.82

0.57

UNet

0.78

0.68

0.91

0.73

0.88

0.59

Res-UNet

0.81

0.67

0.94

0.73

0.90

0.61

spore-UNet

0.82

0.81

0.98

0.80

0.96

0.71

  1. This table displays evaluation metrics measuring precision, sensitivity, specificity, F1-score, accuracy, and IOU. Spore-UNet demonstrates improved performance across all metrics compared to Cellpose, Omnipose, baseline UNet and Res-UNet. Largest values for each column are in bold.
Back to article page