Deep Texture and Structure Aware Filtering Network for Image Smoothing

Abstract

Image smoothing is a fundamental task in computer vision, that attempts to retain salient structures and remove insignificant textures. In this paper, we aim to address the fundamental shortcomings of existing image smoothing methods, which cannot properly distinguish textures and structures with similar low-level appearance. While deep learning approaches have started to explore structure preservation through image smoothing, existing work does not yet properly address textures. To this end, we generate a large dataset by blending natural textures with clean structure-only images, and use this to build a texture prediction network (TPN) that predicts the location and magnitude of textures. We then combine the TPN with a semantic structure prediction network (SPN) so that the final texture and structure aware filtering network (TSAFN) is able to identify the textures to remove (“texture-awareness”) and the structures to preserve (“structure-awareness”). The proposed model is easy to understand and implement, and shows good performance on real images in the wild as well as our generated dataset.