Abstract
The refinement of existing CNN-based Super-resolution Reconstruction (SR) networks mainly focuses on deeper network architecture, which hinders the transmission of information in the networks. A deeper network is unable to make full use of intermediate correlation features and makes the training of the network difficult. To solve these problems, we propose a multi-scale channel attention residual network (MCAR). Specifically, we propose a multi-scale channel attention fusion module (MCAF) to learn local and global channels feature and capture the long-range dependencies. Furthermore, the multi-scale block is adopted to get the different scale feature representations. The experimental results on four benchmark datasets demonstrate that our models can effectively improve the visual effect of images, and outperform most of the advanced SISR methods in PSNR and SSIM.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.