search

cyclomon / UNSB

Official Repository of "Unpaired Image-to-Image Translation via Neural Schrödinger Bridge" (ICLR 2024)
MIT License
154 stars 5 forks source link

Translation direction at inference #19

Closed nickk124 closed 5 months ago

nickk124 commented 5 months ago

Hi, When I train a model with the default --direction (AtoB), will using --direction BtoA at inference work properly for translating some domain B image to domain A, or does the model need to instead be trained with --direction BtoA? I.e., is the model trained for both directions?

I was unsure upon noticing that the training only logs realA->fakeB translations.

Thanks!

cyclomon commented 5 months ago

Hi,

Our training scheme is designed for single direction, so if the model is trained on AtoB, we can only translate AtoB direction.

For BtoA inference, we need another model which is trained for BtoA.

Thansk!

nickk124 commented 5 months ago

I thought so, thanks for confirming!