what exact patch_size is used for the result x2 in your paper?

sanghyun-son / EDSR-PyTorch

PyTorch version of the paper 'Enhanced Deep Residual Networks for Single Image Super-Resolution' (CVPRW 2017)

MIT License

2.43k stars 670 forks source link

what exact patch_size is used for the result x2 in your paper? #330

Open c-yn opened 2 years ago

c-yn commented 2 years ago

Hello, your paper says, "For training, we use the RGB input patches of size 48×48 from LR image with the corresponding HR patches." but in your code, you set patch_size of 192. Could you please tell me which is right? or how does patch_size impach the result... Thanks.