Single Image Super-Resolution: From Sparse Coding to Deep Learning

Recently, deep learning has been successfully applied in numerous areas of computer vision, including low-level image restoration problems. For single image super-resolution (SR), which is an ill-posed problem that tries to recover a high-resolution image from its low-resolution observation, a number of models based on deep neural networks have been proposed and obtained superior performance that overshadows all previously handcrafted models. To regularize the solution of the problem, older methods have focused on using good priors from natural images such as sparse representation, or directly learning the priors from a large data set with models such as deep neural networks. In this chapter, we argue that domain expertise represented by the conventional sparse coding model is still valuable, and it can be combined with the key ingredients of deep learning to achieve further improvements. We …