I am using prediction from a neural net as an input to another network in the auto-context algorithm. I concatenate the prediction with images resulting in size: [2, 2, 64, 96, 96]. The second channel is the output of the previous network which has values for different tissue classes [0, 8].
My question is how do I normalize the prediction channel.
N.B. I am normalizing the 3D image(channel 1) in min-max normalization that changes the value to [0, 1]. Do I do the same also for channel 2?
Thanks.