Image Completion of Highly Noisy Images Using Deep Learning

Generating images from a noisy or blurred image is an active research problem in the field of computer vision that aims to regenerate refined images automatically in a content-aware manner. Various approaches have been developed by academia and industry which includes modern ones applying convolution neural networks and many other approaches to have more realistic images. In this paper, we present a novel approach that leverages the combination of capsule networks and generative adversarial networks that is used for generating images in a more realistic manner. This approach tries to generate images which are locally and globally consistent in nature by leveraging the low level and high-level features of an image.