cszn / SRMD

Learning a Single Convolutional Super-Resolution Network for Multiple Degradations (CVPR, 2018) (Matlab)

Home Page:http://openaccess.thecvf.com/content_cvpr_2018/papers/Zhang_Learning_a_Single_CVPR_2018_paper.pdf

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Grid search strategy

lukk47 opened this issue · comments

For the experiments on real images, a grid search strategy is used as the degradation info. Would you please help provide more details? Thanks

You can use for-loop to produce multiple outputs and choose the best one. Alternatively, you should estimate the blur kernel first.

What's the definition of "the best one"?