About training code

[visionbike]

Hello,

Would you please publish the training code and training code? or Let me know where to find them?

I am considering reconstruct the training stage.

Thank you.

[gfacciol]

This doesn't solve the issue of trining. But if you're more into pytorch then you can use this fork

  https://github.com/gfacciol/CDM-CNN_pytorch

which was used to reproduce the results of CDM-CNN in pytorch for this publication

   T. Ehret, and G. Facciolo, "A Study of Two CNN Demosaicking Algorithms", 
   Image Processing On Line, 9 (2019), pp. 220–230. https://doi.org/10.5201/ipol.2019.274

If you need an example of how to train you can use this code as base:

   https://github.com/gfacciol/demosaicnet_torch/blob/master/train.py

[visionbike]

Thank you so much

[visionbike]

Hi @gfacciol,
I tried your suggesting training code. However, I don't know how to load trained model to test my images. I saw the code in

https://github.com/gfacciol/demosaicnet_torch/blob/master/demosaicnet.py

that just load pretrain model to process image not trained model from

https://github.com/gfacciol/demosaicnet_torch/blob/master/train.py

Thank you.

[gfacciol]

I think that you just need to do `model = torch.load('modelfilename.pth')` to load the weights you trained

…

On Fri, Sep 20, 2019 at 10:01 AM Nguyen Thanh Thien Phuc < ***@***.***> wrote: Hi @gfacciol <https://github.com/gfacciol>, I tried your suggesting training code. However, I don't know how to load trained model to test my images. I saw the code in https://github.com/gfacciol/demosaicnet_torch/blob/master/demosaicnet.py that just load pretrain model to process image not trained model from https://github.com/gfacciol/demosaicnet_torch/blob/master/train.py Thank you. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1?email_source=notifications&email_token=ABGFI2QDSHYZVJNTIG4SFCLQKR7ORA5CNFSM4ITYWFW2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD7F5OJA#issuecomment-533452580>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABGFI2QIYU2VONCTCCLXEJDQKR7ORANCNFSM4ITYWFWQ> .