Training details for EDSR(x1) ? about meta-sr-pytorch

[XuecaiHu]

1. yes. bicubic down and then upscaling to get the training pairs.
2. no, you should change the settings in option.py by following the EDSR paper
3. same as our Meta-RDN

when I have time, i will add the training and test code soon.
Thank you for your good question.

[lisosia]

Thank you very much for your answer!
train/test code would be very helpful 👍

[matri123]

Hi, the upscaling process uses the bicubic interpolation or other interpolation methods?

[lisosia]

reopened by mistake