Abstract
The advantages of confocal microscopy over widefield microscopy are their ability to produce optically sectioned images and their ability to produce multi-color imaging in which different organelles within the biological specimen are stained using multiple dyes, enabling colocalization studies. These features make confocal microscopy a widely used tool to provide valuable morphological and functional information within cells and tissues. One of the major drawbacks of confocal microscopy is its limited spatial resolution. Here, we train two generative adversarial networks using paired and unpaired data of low- and high-resolution images to improve the spatial resolution in confocal microscopy.