Pix2Pix-for-Semantic-Segmentation-of-Satellite-Images
Pix2Pix-for-Semantic-Segmentation-of-Satellite-Images copied to clipboard
Converting generator outputs to segmentation masks in discrete space
Hi there,
Thanks for making this notebook public available. I just have one question after reading your code. Your generator basically outputs tanh activations, which are not integers that stand for the classification results for each pixel. How did you process these activations to get the final segmentation masks?
Looking forward to your reply, thanks