search

phoenix104104 / LapSRN

Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution (CVPR 2017)
http://vllab.ucmerced.edu/wlai24/LapSRN/
449 stars 120 forks source link

About the training dataset #18

Closed thecryboy closed 4 years ago

thecryboy commented 5 years ago

thanks for your code.As you mentioned in the paper,the training dataset is t91 and bsd 200.However,the dataset is t91+bsd200+g100... So ,what is the result for your paper? Is is consist of General 100?

nachifur commented 4 years ago

I have the same problem and look forward to your reply.

phoenix104104 commented 4 years ago

The results in our paper are using the training data from T91 + BSB200 only. You are welcome to use G100 for further study.