Compression Artifacts Removal Using Convolutional Neural Networks

Abstract

This paper shows that it is possible to train large and deep convolutional neural networks (CNN) for JPEG compression\nartifacts reduction, and that such networks can provide significantly better reconstruction quality compared to\npreviously used smaller networks as well as to any other state-of-the-art methods. We were able to train networks\nwith 8 layers in a single step and in relatively short time by combining residual learning, skip architecture, and\nsymmetric weight initialization. We provide further insights into convolution networks for JPEG artifact reduction\nby evaluating three different objectives, generalization with respect to training dataset size, and generalization with\nrespect to JPEG quality level.