Coherent-Image Reconstruction Using Convolutional Neural Networks

Abstract

We describe a technique for incorporating convolutional-neural-network models into a comprehensive approach for coherent-image reconstruction in the presence of noise and phase errors using the consensus equilibrium framework.