fVDB-Reality-Capture is a reality-capture toolbox built on top of fVDB. It provides high-level abstractions and APIs for common reality capture tasks, such as loading sensor data, reconstructing radiance fields, learning scale-conditioned instance feature fields, extracting meshes and point clouds, visualization, and exporting results across standard formats such as PLY and USD (USDC/USDZ).
By leveraging the power of fVDB, fVDB-Reality-Capture can scale reconstruction to very large or dense inputs, while maintaining high performance and low memory usage. fVDB has 50% better throughput than gsplat in end-to-end training benchmarks and 30% lower runtime, while producing higher quality results and working out-of-the-box on a wide range of inputs. The videos below show large-scale reconstructions of complex scenes using fVDB-Reality-Capture.
For more information about what fVDB-Reality-Capture can do, tutorials and documentation, please see the fVDB-Reality-Capture documentation.
The fvdb-reality-capture Python package can be installed using published packages with pip or built from source.
For the most up-to-date information on installing fVDB-Reality-Capture's pip packages, please see the installation documentation.
Note: For more complete instructions for building
fvdb-corefrom source, including setting up a build environment and obtaining the necessary dependencies, see the fVDB README.
Clone the fvdb-core repository and the fvdb-reality-capture repository.
git clone git@github.com:openvdb/fvdb-core.git
git clone git@github.com:openvdb/fvdb-reality-capture.gitNext, build and install the fVDB library
pushd fvdb-core
./build.sh install verbose
popdFinally, install fVDB-Reality-Capture
pushd fvdb-reality-capture
pip install .
popdfVDB and fVDB-Reality-Capture were first developed by the NVIDIA High-Fidelity Physics Research Group within the NVIDIA Spatial Intelligence Lab, and continues to be developed with the OpenVDB community to suit the growing needs for a robust framework for spatial intelligence research and applications.
fVDB-Reality-Capture is built on top of fVDB, which provides efficient GPU data structures and algorithms for working with sparse volumetric data. By leveraging the power of fVDB, fVDB-Reality-Capture can scale reconstruction to very large or dense inputs, while maintaining high performance and low memory usage.
fVDB and fVDB-Reality-Capture are open source under the Apache 2.0 license. We welcome contributions and feedback from the community.
For questions or feedback, please use the GitHub Issues for this repository.