A Two-Stage Three-Dimensional Attention Network for Lightweight Image Super-Resolution

Lei CHEN, Yanjie YANG, Xu ZHUANG, Jason WANG, Qin MAO*, Hong YUE, Xuekai WEI, Fei CHENG, Xuemei ZONG, Mingliang ZHOU

*Corresponding author for this work

Research output: Journal PublicationsJournal Article (refereed)peer-review

Abstract

In recent years, single image super-resolution (SISR) methods using convolutional neural networks (CNN) have achieved satisfactory performance. Nevertheless, the large model scale and the slow inference speed of these methods greatly limit the application scenarios. In this paper, we propose a two-stage three-dimensional attention network (ATTNet) for lightweight image super-resolution. First, we put forward the spatial feature encoder–decoder (SFE-D) with a spatial attention mechanism. Next, the channel transposed attention module (CTAM) with a channel self-attention mechanism is designed. Both the modules are used for fine feature extraction in the low-resolution stage. Finally, the content-based pixel recombination module (CPRM) is proposed to reconstruct the detailed content with a joint attention mechanism in the high-resolution stage. According to our experimental results, significant performance in terms of the quantitative metrics and the subjective visual quality can be achieved on average compared with the state-of-the-art lightweight SISR algorithms.

Original languageEnglish
Article number2354017
JournalInternational Journal of Pattern Recognition and Artificial Intelligence
Volume37
Issue number13
DOIs
Publication statusPublished - 1 Oct 2023
Externally publishedYes

Bibliographical note

Publisher Copyright:
World Scientific Publishing Company.

Keywords

  • attention mechanism
  • high-resolution stage
  • lightweight
  • low-resolution stage
  • Super-resolution

Fingerprint

Dive into the research topics of 'A Two-Stage Three-Dimensional Attention Network for Lightweight Image Super-Resolution'. Together they form a unique fingerprint.

Cite this