Understanding the whole process

[vsantjr]

Hello, First of all, congrats for your interesting work. I am just trying to figure out how is the entire process for using RCAN. I have a small dataset for training with low resolution images (128 x 128). I want to use RCAN to increase the resolution of the images to 1024 x 1024. Hence, this is what I could understand:

For training:

1. Place the original training set in 'OriginalTestData'.
2. Run 'Prepare_TestData_HR_LR.m' in Matlab to generate HR/LR images with different degradation models.
3. Run (input=128x128, output=1024x1024) python main.py --model RCAN --save my_name --scale 8 --n_resgroups 10 --n_resblocks 20 --n_feats 64 --reset --chop --save_results --print_model --patch_size 1024 --pre_train ../experiment/model/RCAN_BIX8.pt

For inference/testing (to generate the high resolution images)

Steps 1 and 2 before Plus the step below:

python main.py --data_test MyImage --scale 8 --model RCAN --n_resgroups 10 --n_resblocks 20 --n_feats 64 --pre_train ../model/RCAN_BIX8.pt --test_only --save_results --chop --save 'RCAN' --testpath ../LR/LRBI --testset my_test_set

Moreover, I understood that I must split my original training dataset into two: one for training/validation and another for testing. Is it?

Thank you.