二值图像超分辨率重建网络  

Binary neural networks for image super-resolution

作　　者：姜馨蕊 王楠楠[1] 辛经纬 李柯宇 杨曦 高新波 Xinrui JIANG;Nannan WANG;Jingwei XIN;Keyu LI;Xi YANG;Xinbo GAO(State Key Laboratory of Integrated Services Networks,Xidian University,Xi'an 710071,China;Chongqing Key Laboratory of Image Cognition,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区：[1]西安电子科技大学综合业务网理论及关键技术国家重点实验室,西安710071 [2]重庆邮电大学重庆市图像认知重点实验室,重庆400065

出　　处：《中国科学：信息科学》2021年第10期1690-1705,共16页Scientia Sinica(Informationis)

基　　金：2030新一代人工智能重大专项(批准号:2018AAA0103202);国家自然科学基金(批准号:61922066,61876142,62036007)资助项目。

摘　　要：近年来,深层卷积神经网络在图像超分辨率重建任务中取得了巨大成功,然而复杂的深度神经网络会消耗大量存储空间以及计算资源,严重限制了其在资源有限的移动端设备上的部署.因此降低模型的资源消耗将有助于扩展深度超分辨率网络的实际应用范围.二值神经网络占用存储空间小、计算效率高,激励我们将二值化算法应用于目前的深度超分辨率重建领域,满足现有移动设备对于超分辨率的实际应用需求.因此,本文关注于二值图像超分辨率重建网络的研究.为此,我们首先总结了现有二值化方法,并针对其技术细节和算法特点进行了详细介绍.随后,我们探索了目前二值化方法在超分辨率领域的实际应用效果,并面向图像超分辨率重建任务提出一种新的二值化算法,主要通过提高网络前向过程表达能力和减少网络反向过程训练损失提升二值超分辨率网络的性能.实验表明,无论对比现有基于分类任务的二值化算法还是对比基于超分辨率任务的二值化算法,我们的方法均可以取得最优的性能.Recently, deep convolutional neural networks(DNNs) have achieved state-of-the-art performance on the super-resolution task. However, bigger and deeper networks often lead to high computational cost and high memory usage, preventing massive applications on resource-limited devices. Therefore, reducing the model storage and computation costs will expand the application range of DNNs. Binary neural networks, which could reduce the model size and allow for efficient inference, are energy-efficient for embedded devices. It motivates us to apply the binary methods into super-resolution field, enabling us to meet the requirements of the hardware platforms in practical applications. Therefore, this paper focuses on the binary image super-resolution network.To this end, we first review the current binary neural networks and introduce the technical details and algorithm characteristics of the existing algorithms in detail. Then, we propose to apply the existing binary training methods to the super-resolution task. On this basis, we propose a new binary super-resolution method, which advances network performance by improving the representation ability of the quantized network in the forward propagation and reducing information loss in the backward propagation. Extensive experiments demonstrate that the proposed method not only outperforms existing binary methods but also outperforms the state-of-the-arts binary super-resolution methods.

关 键 词：二值卷积神经网络 图像超分辨率重建 二值化 量化 模型压缩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]