Table 2.

Comparative evaluation of our U-Net model against U-Net-base, Deeplabv3, and Seg-Net in terms of performance metrics relevant to bacterial enumeration.

Net	DSC	Recall	Precision	F1
U-Net-base	0.685 ± 0.034	0.777 ± 0.052	0.766 ± 0.057	0.757 ± 0.033
Deeplabv3	0.579 ± 0.150	0.695 ± 0.066	0.816 ± 0.105	0.734 ± 0.049
Seg-Net	0.663 ± 0.045	0.717 ± 0.063	0.815 ± 0.031	0.747 ± 0.042
Our U-Net	0.688 ± 0.042	0.787 ± 0.046	0.772 ± 0.029	0.762 ± 0.029