I am bit confused about Discriminator


How are you utilizing the Discriminator in your work? Does it receive randomly rendered outputs for fake signal and real images for real signal, or does it receive the same position rendered output each time for fake signal?

Yes, we’re utilizing the discriminator in GET3D. At every training iteration, it receives a rendered image (rendered at a random camera position, which is sampled from the training dataset) and classifies whether this image is real or fake. The real and fake images are not rendered in the same camera position, all of the images & cameras are randomly sampled from the training set

1 Like

Thank You !

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.