Would it be possible to get the camera parameters that are respresented via their prompt tags?
#6
by SDuser23487 - opened
Were the camera positions during training picked at random locations or fixed? If they were fixed would it be possible to get the camera data the LoRA was trained on?
That way we could reconstruct the position and generate a ControlNet image for Qwen in a 3D app from a coarse 3D model.
Great LoRA,
Thank you so much :)
SDuser23487 changed discussion status to closed