CasaPuNet: Channel Affine Self-Attention- Based Progressively Updated Network for Real Image Denoising

Abstract

Recently, the popularity of deep learning has brought broad applications of computer vision technology in industrial information systems. However, the process of image acquisition will inevitably introduce noise, which may heavily degrade image visual quality. Most of the proposed denoising methods are non-blind and they have limited performance in removing real noise with different noise levels. To overcome this problem, we propose a deep CNN-based blind model, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , channel affine self-attention based progressively updated network (CasaPuNet) for real image denoising. First, we introduce degradation mapping module (DMM) to extract degradation information, which makes the remaining subnetwork of CasaPuNet perform non-blind denoising. Then, CasaPuNet adopts a multi-stage architecture, which resolves the large gap between the noisy input and clean output into several small gaps and eliminates these small gaps step by step through progressive inference. Finally, a novel channel affine self-attention (CASA) is designed to adaptively fuse the features from multiple stages according to input statistics. Specifically, CASA extracts channel information from different features and converts them into channel weights through an affine structure for adaptive adjustment. CASA brings a significant performance gain with a small number of parameters. Extensive experiments demonstrate that CasaPuNet outperforms state-of-the-art denoising methods both quantitatively and visually. Code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/chaoren88/CasaPuNet</uri> .