About the loss function

[DylanWusee]

trafficstars

Hi Xingyu,

Thank you for releasing code here.

I noticed that the loss function used in the code is different from the one in the Paper. In the paper, you used a huber loss together with a cycle-consistency regularization.

Are there particular reasons that you removed it from your code?

Thank you very much. Best, Dylan

[xingyul]

We provided huber loss in the code.

[DylanWusee]

But the huber loss in your code doesn't have the cycle-consistency regularization term. Is that correct?

[xingyul]

That's correct. Cycle-consistency regularization should be easy to implement. Do you want me to upload the implementation?

[HimangiM]

I am having some issues in implementing it. It would be great if you can upload it.