Asymmetric CNN for Image Superresolution

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review

8 Scopus Citations
View graph of relations


  • Chunwei Tian
  • Yong Xu
  • Wangmeng Zuo
  • Chia-Wen Lin
  • David Zhang

Related Research Unit(s)


Original languageEnglish
Pages (from-to)3718-3730
Journal / PublicationIEEE Transactions on Systems, Man, and Cybernetics: Systems
Issue number6
Online published3 May 2021
Publication statusPublished - Jun 2022


Deep convolutional neural networks (CNNs) have been widely applied for low-level vision over the past five years. According to the nature of different applications, designing appropriate CNN architectures is developed. However, customized architectures gather different features via treating all pixel points as equal to improve the performance of given application, which ignores the effects of local power pixel points and results in low training efficiency. In this article, we propose an asymmetric CNN (ACNet) comprising an asymmetric block (AB), a memory enhancement block (MEB), and a high-frequency feature enhancement block (HFFEB) for image superresolution (SR). The AB utilizes one-dimensional (1-D) asymmetric convolutions to intensify the square convolution kernels in horizontal and vertical directions for promoting the influences of local salient features for single image SR (SISR). The MEB fuses all hierarchical low-frequency features from AB via a residual learning technique to resolve the long-term dependency problem and transforms obtained low-frequency features into high-frequency features. The HFFEB exploits low- and high-frequency features to obtain more robust SR features and address the excessive feature enhancement problem. Additionally, it also takes charge of reconstructing a high-resolution image. Extensive experiments show that our ACNet can effectively address SISR, blind SISR, and blind SISR of blind noise problems. The code of the ACNet is shown at

Research Area(s)

  • Asymmetric architecture, blind single image superresolution (SISR), Convolution, convolutional neural network (CNN), Degradation, Feature extraction, image superresolution (SR), Kernel, multilevel feature fusion, multiple degradation task, Superresolution, Task analysis, Training

Citation Format(s)

Asymmetric CNN for Image Superresolution. / Tian, Chunwei; Xu, Yong; Zuo, Wangmeng; Lin, Chia-Wen; Zhang, David.

In: IEEE Transactions on Systems, Man, and Cybernetics: Systems, Vol. 52, No. 6, 06.2022, p. 3718-3730.

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review