Fig. 5: Segmentation examples for LIVECell dataset using LACSS models trained with the semi-supervision configuration.
From: Efficient end-to-end learning for cell segmentation with machine generated weak annotations
Examples for all eight cell lines were shown. The input phase contrast images (left) were overlaid with model predictions drawn as yellow contours. The ground truth segmentations were shown on the right, for which the model predictions were also shown as white contours for comparison. Scale bars represent 20 micrometers.
