I am trying to train my model for semantic segmentation. Here in my images in datastore are 3000 by 5000 pixels. but the input layer of the network is 32 by 32. I used augmenteddatastore mommand for resizeing the image datastore. Later on to i need to combine the pixellabel database and reshaped image datastore. but the comman
doesnot accept the augmented datastore. so i need to either convert augmenteddatastore to image datastore or any other method to combine the training data.