Porous media flow estimation using Deep Convolutional Networks Implementation of a Deep Autoencoder network to predict single-phase flow in porous media from simulation data generated by FEniCs library.