search

qibao77 / CFSNet

pytorch code of "CFSNet: Toward a Controllable Feature Space for Image Restoration"（ICCV2019）
62 stars 4 forks source link

input/output channel #4

Open sjf18 opened 4 years ago

sjf18 commented 4 years ago

hi, thanks for your great, when i'm using your trained models, i have some questions. 1). why the input and output channel in deblock and denoise is 1 ? are those two models only support gray images? 2). i found the model for 4✖️sr could not be load properly

qibao77 commented 4 years ago

I am sorry for the late reply. 1) For JPEG image deblocking task, we set the input and output channels to 1 for fair comparison with other methods. You can also set the input and output channels to 3 to support RGB images. In fact, we provide color image denoising results in our appendix. 2) You can check whether the path of the saved model is correct.