Published October 5, 2021 | Version v1
Software Open

Pre-trained Pytorch Models used for IEEE Signal Processing Letters paper titled 'Improving Super-Resolution Performance using Meta-Attention Layers'

Authors/Creators

  • 1. University of Edinburgh/University of Malta
  • 2. University of Malta

Description

This collection contains all the pre-trained model weights used to produce the results in our paper titled 'Improving Super-Resolution Performance using Meta-Attention Layers' (link here).  Instructions on how to use these models is provided in our Github repo here.  

Notes

These models were produced as part of the Deep-FIR project, which is financed by the Malta Council for Science & Technology (MCST) (grant number R&I-2017-002-T), for and on behalf of the Foundation for Science & Technology, through the FUSION: R&I Technology Development Programme.

Files

models_blurry_compressed_images.zip

Files (3.3 GB)

Name Size
md5:f810a8954ce67b20d6c6d9a2676205af
704.1 MB Preview Download
md5:26a095183722f3dbfd616df4ca545b83
2.0 GB Preview Download
md5:eb2afece8b40081aa74e8f5e2cd98673
583.1 MB Preview Download

Additional details

Related works

Is supplement to
Journal article: 10.1109/LSP.2021.3116518 (DOI)