Skip to main content

Table 1 Performance (cell identification error) benchmarking of DelTA trained with SyMBac. Models trained with SyMBac perform well for both oil and air objectives and outperform models trained with user-generated training data

From: Synthetic Micrographs of Bacteria (SyMBac) allows accurate segmentation of bacterial cells using deep neural networks

Dataset

DeLTA (pretrained)

DeLTA (SyMBac trained)

Omnipose (SyMBac trained)

100×, oil, NA 1.45, [30]

0.93%

1.1%

0.25%

100×, oil, NA 1.49 [3]

1.8%

0.47%

0.12%

60×, air, NA 0.95 [3]

N/A

0.35%

0.28%

Back to article page

BMC Biology

ISSN: 1741-7007

Contact us