Make it condition GAN? And with FNO.

[Tariq97]

Hello,

Many thanks for this nicely written code.
I would like to know if we can make his architecture as a conditioned GAN? By embedding more channels(of condition parameters) to the inputs?

Also, do you think if we can use the conventional FNO instead of UNO? How do you think the results are going to change in that case?

Looking forward to hearing from you.

Thanks

[kazizzad]

Hi, yes to both of your questions.
You can make it conditional GANO the way you said it.
You can use any neural operator architecture instead of UNO. GANO framework is independent of the choice of neural operator architecture.