The goal is to recover a high resolution image from a low resolution input using deep convolutional neural networks