Hello! I am trying to start 3d reconstruction with your 3DGS software.
I'd like to ask some basic questions.
Does 3DGS needs camera intrinsic parameter? It just helps 3D quality, or it is must to have?
I am using Colmap to make SfM as prerequisite before train.py used to make 3DGS .ply file. I put 2800 images using basic gpu 16GB VRAM (10 fps from a movie), then takes nearly 30 hours to make a .ply for 3DGS. Do you know any good way to speed it up? like other tool.
so I should put .ply to train.py, right? or what do you mean "Colmap dataset" in readme?
Thank you for your kind advice!
Hello! I am trying to start 3d reconstruction with your 3DGS software. I'd like to ask some basic questions.