Skip to main content
Fig. 5 | PhotoniX

Fig. 5

From: Enhancing image resolution of confocal fluorescence microscopy with deep learning

Fig. 5

(a) Diffraction-limited confocal microscopy images with different pixel sizes are used as input to the network. (b) The Super-resolved network inference images using L1 loss before upsampling/downsampling the input images to match the pixel size of the training data. (c) The Super-resolved network inference images using L2 loss before upsampling/downsampling the input images to match the pixel size of the training data. (d) The super-resolved network inference images using L2 loss after upsampling/downsampling the input images to match the pixel sizes of the training data. (e) The STED images of the same field of view. Scale bars in (e) are 8 μm, 6 μm, 4 μm, 3 μm, 3 μm, 3 μm and 2 μm, respectively from the first row to the seventh row

Back to article page