Is there possible way to generate novel view conditioned on input images?

[williamium3000]

Thanks for the nice work. Does the method support generating novel view conditioned on input images?

[universome]

Hi! No, there is not such possibility. One way to achieve this would be embedding an image into the latent space of the generator and then synthesizing novel views. However, this functionality is not supported in the repo, and we have not tested how well it would work.

[williamium3000]

Thanks, I will try to see if it works some time later