This is an nerfstudio framework based implementation for NeRFPlayer.
NeRFPlayer follows the integration guidelines described here for custom methods within nerfstudio.
Follow these instructions up to and including "tinycudann" to install dependencies and create an environment
git clone https://github.com/lsongx/nerfplayer-nerfstudio.git
Navigate to this folder and run python -m pip install -e .
Run ns-train -h
: you should see a list of "subcommands" with nerfplayer-nerfacto
and nerfplayer-ngp
included among them.
Now that NeRFPlayer is installed you can play with it.
- Currently only DyCheck dataset is supported.
- Download DyCheck data.
- You can capture your own scenes by following DyCheck's guide.
- Launch training with
ns-train nerfplayer-ngp --data <data_folder>
. This specifies a data folder to use.- example:
ns-train nerfplayer-ngp --data dycheck/mochi-high-five/
- example:
- Connect to the viewer by forwarding the viewer port, and click the link to
viewer.nerf.studio
provided in the output of the train script
Please open Github issues (under this repo, not under nerfstudio) for any installation/usage problems you run into.
- Multi-camera datasets: DyNeRF, ImmersiveVideo
- Decomposition in NeRFPlayer. Under nerfstudio's framework, we got NaN soon if a decomposition module is used.