Skip to main content
. 2020 Sep 29;25(9):096009. doi: 10.1117/1.JBO.25.9.096009

Fig. 3.

Fig. 3

Deep-learning-based fluorescence channel AF. AF model is tested on U2OS-AF dataset. (a) The full FOV of a test image at different z depths, (b) the zoom-in areas as the inputs, (c) the zoom-in area prediction, and (d)–(g) the different zoom-in areas (marked by colored boundaries) of input and prediction, respectively. The ground truth column is put on the right with corresponding views of input and prediction for comparison.