Machine and deep-learning techniques for image super-resolution

Abstract

Super-resolution is a class of techniques for upscaling images or video. Owing to the significant advancements in the field of machine learning and deep learning in the last decade, this is an active problem with a plethora of approaches proposed. This chapter aims at providing a basic understanding of the traditional as well as machine-learning and deep-learning-based approaches used in image super-resolution. We also discuss in depth the various loss functions used for training the neural networks and the performance metrics used for evaluating them. Finally, we conclude the chapter by providing a comparative study of some recent popular approaches to super-resolution and insights into future research directions. © 2023 The Institute of Electrical and Electronics Engineers, Inc.