Unpaired Screen-Shot Image Demoiréing with Cyclic Moiré Learning

Abstract

We propose an end-to-end unpaired learning approach to screen-shot image demoireing based on cyclic moire learning. The proposed cyclic moire learning algorithm consists of the moireing network and the demoireing network. The moireing network generates moire images to construct a pseudo-paired set of moire and clean images. Then, the demoireing network is trained in a supervised manner using the generated pseudo-paired dataset to remove moire artifacts. In the moireing network, the moire generation is separately learned as global pixel intensity degradation and moire pattern generation for more realistic moire artifact generation. Furthermore, the moireing network and the demoireing network are integrated together to be trained in an end-to-end manner. Experimental results on different datasets demonstrate that the proposed algorithm significantly outperforms state-of-the-art unsupervised demoireing algorithms as well as image restoration algorithms.