You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for releasing the code!
I would like to know how to control the output image to be in those six viewpoints without showing the input camera position.
The text was updated successfully, but these errors were encountered:
Thank you for your reply!
I want to know how you control the output image to be the fixed six viewing angles without generating images outside the six viewing angles.
In other issue I see you do not explicitly use any camera pose input during training or inference, but I'm not quite sure how you control the synthesis of the novel views from those six specific angles.
I have another question: why does the decoding of latents directly result in one large image that includes six novel views? I don’t quite understand how this is achieved. I look forward to your reply, thank you!
Thanks for releasing the code!
I would like to know how to control the output image to be in those six viewpoints without showing the input camera position.
The text was updated successfully, but these errors were encountered: