i trained segmentation net, and predicted a mask on a new file.
now i want to show the result (predicted mask on the original image), but here is the problem.
the original image have another size then the net input’s size, and hence have another size then the output mask. since in the prediction stage fastai automatically resize the input image, when i write
fastai crops the input image instead of resize it, anf i get the input image cropped to the mask size, and the mask plotted on the left-top corner of the input image…
how can i show it correctly?