I'd like to use the method on a case where the object rotates around itself and the camera is at a fixed position. Could I tell the reconstruction pipeline to somehow account for this? Has anyone tried this? I could not find any implementation.
In theory this may work if you had a solid background colour and well-lit lighting. The NeRF will try to reconstruct the background anyways and probably will be messy tho
I'd like to use the method on a case where the object rotates around itself and the camera is at a fixed position. Could I tell the reconstruction pipeline to somehow account for this? Has anyone tried this? I could not find any implementation.