NTIRE 2024 Challenge on Efficient Super-Resolution @ CVPR 2024

LSANet: Efficient Image Super-Resolution with Lightweight Spaced Attention Mechanism

Source code, pretrained model and fact sheet of our solution to NTIRE 2024 Efficient Super-resolution. More details can be found in the official repository of the challenge https://github.com/Amazingren/NTIRE2024_ESR

The Environments

The evaluation environments adopted by us is recorded in the requirements.txt. After you built your own basic Python setup via either virtual environment or anaconda, please try to keep similar to it via:

pip install -r requirements.txt

or take it as a reference based on your original environments.

The Validation datasets

After downloaded all the necessary validate dataset (DIV2K_LSDIR_valid_LR and DIV2K_LSDIR_valid_HR), please organize them as follows:

|NTIRE2024_ESR_Challenge/
|--DIV2K_LSDIR_valid_HR/
|    |--000001.png
|    |--000002.png
|    |--...
|    |--000100.png
|    |--0801.png
|    |--0802.png
|    |--...
|    |--0900.png
|--DIV2K_LSDIR_valid_LR/
|    |--000001x4.png
|    |--000002x4.png
|    |--...
|    |--000100x4.png
|    |--0801x4.png
|    |--0802x4.png
|    |--...
|    |--0900.png
|--NTIRE2024_ESR/
|    |--...
|    |--test_demo.py
|    |--...
|--results/
|--......

How to test the model?

1. https://github.com/BhJia/NTIRE2024_ESR
2. Select the model you would like to test from run.sh

   CUDA_VISIBLE_DEVICES=0 python test_demo.py --data_dir [path to your data dir] --save_dir [path to your save dir] --model_id 44

   • Be sure the change the directories --data_dir and --save_dir.
3. More detailed example-command can be found in run.sh for your convenience.

About Fact Sheet

All fact sheet files are in folder Fact Sheet.

team44_Fact_Sheet.pdf is the compiled pdf file of our fact sheet. team44_Fact_Sheet.zip is the corresponding .tex source files. Important files in zip file are EfficientSR_factsheet.tex which is .tex source file. egbib.bib is the BibTex form reference file. fig.pdf and fig1.pdf are source images of the fact sheet. You can select zip file and upload it to a common LaTeX editor to check the details.

License and Acknowledgement

This code repository is release under MIT License.