Open yinyiyu opened 2 years ago
The low resolution image is preprocessed by bicubic interpolation and then fed into the network. That is why the input and output of the model have same size.
Thank you for your answer. I also noticed that bilinear interpolation was first used as the input of the model. One more question, why the input and output are gray picture instead of RGB graphs?
because we evaluate PSNR on gray channel only to compare with other methods.
Hello, thank you for your excellent work. I noticed that in your code, the input and output of the model are all of the same size. Isn't this super resolution work? Why are the input and output the same?