representation

Issue I’m looking for a neural network model that can extract useful information from an image. Here "useful" is arbitrarily