Title:Multi-scale Dynamic Feature Encoding Network for Image Demoireing

Abstract:The prevalence of digital sensors, such as digital cameras and mobile phones, simplifies the acquisition of photos. Digital sensors, however, suffer from producing Moire when photographing objects having complex textures, which deteriorates the quality of photos. Moire spreads across various frequency bands of images and is a dynamic texture with varying colors and shapes, which pose two main challenges in demoireing---an important task in image restoration. In this paper, towards addressing the first challenge, we design a multi-scale network to process images at different spatial resolutions, obtaining features in different frequency bands, and thus our method can jointly remove moire in different frequency bands. Towards solving the second challenge, we propose a dynamic feature encoding module (DFE), embedded in each scale, for dynamic texture. Moire pattern can be eliminated more effectively via this http URL proposed method, termed Multi-scale convolutional network with Dynamic feature encoding for image DeMoireing (MDDM), can outperform the state of the arts in fidelity as well as perceptual on benchmarks.