The implementation of "LCRCA: Image Super-Resolution Using Lightweight Concatenated Residual Channel Attention Networks"
Dependencies： pytorch1.2.0 or 1.1.0 opencv-python（installed using pip） scikit-learn scikit-image imageio tqdm
Train： 1. open a terminal and cd to src 2. type and run: python main.py --model (model name） --scale scale --batch_size batch_size --patch_size patch_size --save "name_of_file_foder" 3. data will be saved to: ../experiment/"name_of_file_foder"
Test： 1. 1. type and run: python --main.py --model xxx --scale x --data_test Set5+Set14+B100+Urban100 --load ../experiment/"name_of_file_foder"/model/model_best.pt --test_only --self_ensemble --save_results --save --self_ensemble
file description： src/model/grnn.py——Network of LCRCA option.py——Configuration file
Thanks sanghyun-son et.al. for their excellent open source project at: https://github.com/sanghyun-son/EDSR-PyTorch.