Overview Videos
SLIM-VDB Overview
Abstract
This paper introduces SLIM-VDB, a new lightweight semantic mapping system with probabilistic semantic fusion for closed-set or open-set dictionaries. Advances in data structures from the computer graphics community, such as OpenVDB, have demonstrated significantly improved computational and memory efficiency in volumetric scene representation. Although OpenVDB has been used for geometric mapping in robotics applications, semantic mapping for scene understanding with OpenVDB remains unexplored. In addition, existing semantic mapping systems lack support for integrating both fixed-category and open-language label predictions within a single framework. In this paper, we propose a novel 3D semantic mapping system that leverages the OpenVDB data structure and integrates a unified Bayesian update framework for both closed- and open-set semantic fusion. Our proposed framework, SLIM-VDB, achieves significant reduction in both memory and integration times compared to current state-of-the-art semantic mapping approaches, while maintaining comparable mapping accuracy. An open-source C++ codebase with a Python interface is available at https://github.com/umfieldrobotics/slim-vdb.
Approach
SLIM-VDB takes advantage of the highly optimized 3D volumetric representation presented in OpenVDB, a library used in the computer graphics community. Our work integrates semantic fusion at a voxel level with OpenVDB, resulting in a lightweight mapping library that can handle both open-set and closed-set semantics. The key component is Bayesian semantic fusion: this work takes advantage of Dirichlet-Categorical conjugacy and Normal-Normal Inverse Gamma conjugacy to tractably handle different semantic predictions from a network.
Our key contributions are:
- A novel framework that builds on OpenVDB to enable lightweight, memory-efficient semantic mapping.
- A unified Bayesian inference framework that enables either closed-set or open-set semantic estimation.
- An open-source C++ library with a Python interface for easy integration with robotics applications.
Citation
This project was developed in the Field Robotics Group at the University of Michigan - Ann Arbor.
@article{sheppard2026slimvdb,
author = {Sheppard, Anja and Ewen, Parker and Wilson, Joey and Sethuraman, Advaith V. and Adewole, Benard and Li, Anran and Chen, Yuzhen and Vasudevan, Ram and Skinner, Katherine A.},
title = {SLIM-VDB: A Real-Time 3D Probabilistic Semantic Mapping Framework},
journal = {IEEE Robotics Automation and Letters},
volume = {TBD},
year = {2026},
number = {TBD},
article-number = {TBD},
url = {https://ieeexplore.ieee.org/document/11344775},
doi = {10.1109/LRA.2026.3652875}
}