Table 3 Benchmarking LACSS models trained with A431 image data and annotations.

From: Efficient end-to-end learning for cell segmentation with machine generated weak annotations

 

AP50

AP75

mAP

AJI

LACSS

0.84

0.28

0.39

0.65

  1. The training dataset includes 500 immunofluorescence images of A431 cells at high density. All annotations were derived from experimental data semi-programmably. See text for more details.
Back to article page