This is an offical implementation of the CVPR2022's paper [Learning the Degradation Distribution for Blind Image Super-Resolution](https://arxiv.org/abs/2203.04962)
Hi, thanks for the code sharing.
I have one question about the calculating way of noise mean.
Why do you use MSELoss instead of just summing all the noise first and then dividing by the total number?
Using MSELoss will let all single noise values do square first, and then sum up.
It might not get the correct value of the noise mean.
Hi, thanks for the code sharing. I have one question about the calculating way of noise mean. Why do you use MSELoss instead of just summing all the noise first and then dividing by the total number? Using MSELoss will let all single noise values do square first, and then sum up. It might not get the correct value of the noise mean.
Thanks!