Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Exporting Generated Scene #6

Open
karthik101200 opened this issue Jun 17, 2024 · 1 comment
Open

Exporting Generated Scene #6

karthik101200 opened this issue Jun 17, 2024 · 1 comment

Comments

@karthik101200
Copy link

karthik101200 commented Jun 17, 2024

Hello. Thanks for the code. I tried nerfgs for a custom dataset and the results are pretty good. I want to export the scene to a mesh or any other dense representation which can be later imported in a simulation environment. Wanted to ask if there is a way to do so. Thanks.

Also just to add, I didnt run nerfsh first as
give me out of memory error
ns-train nerfsh --data DATA-PATH --pipeline.model.camera-optimizer.mode off

RuntimeError: /tmp/pip-req-build-sf4k0pds/include/tiny-cuda-nn/gpu_memory.h:584 cuMemCreate(&m_handles.back(), n_bytes_to_allocate, &prop, 0) failed: CUDA_ERROR_OUT_OF_MEMORY

Any parameters to tune so that I can train. I have RTX A2000 GPU

@boooolean
Copy link
Collaborator

For exporting the nerfgs, it is currently only directly exportable as a .ply file similar to splatfacto. There are tools you can find online to convert the .ply file into dense representations.

As for nerfsh running out of memory, I am not sure, we will have to look more into it, but similar parameter tunings to standard nerf will also help out. Out of curiosity, how did you run nerfgs without running nerfsh? In our proposed pipeline, we had assumed you first create a nerfsh and then convert it into a nerfgs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants