PolyMerge: Compressing 3D Gaussian Splats with Polytope Coverings for Provably Safe Resource-Constrained Navigation

Jihoon Hong, Chih-Yuan Chiu, Sara Fridovich-Keil, Glen Chou
Georgia Institute of Technology
IEEE Robotics and Automation Letters, 2026
Paper arXiv (coming soon) Code (coming soon)
PolyMerge overview: a 3DGS model is converted into multiple polytope covers offline; the chosen cover is loaded to a Crazyflie microcontroller, which performs online CBF filtering for safe on-board navigation.

Overview. PolyMerge converts a 3DGS model into multiple polytope covers, each over-approximating all obstacles with a different polytope count and level of conservativeness (left). A cover satisfying the hardware-driven compute constraints is loaded to the drone microcontroller (middle), which performs CBF filtering of nominal control inputs to navigate safely, entirely on-board (right).

Abstract

Obstacle avoidance is essential for safe navigation and motion planning. Recent radiance field reconstruction methods enable object detection and modeling with high fidelity, but remain too memory- and compute-intensive for on-board perception-based path planning. To address these limitations, we propose PolyMerge to convert a large, photorealistic 3D Gaussian Splatting (3DGS) model of a scene into a lightweight representation of convex polytopes whose union provably over-approximates all obstacles in the original 3DGS model. PolyMerge tunes the polytope count to trade off conservativeness and compute cost, and integrates with control barrier functions (CBFs) to plan collision-free paths. We showcase PolyMerge in simulation and hardware experiments on a Crazyflie drone, which uses PolyMerge to compute and follow safe trajectories in real time under severe onboard compute resources, outperforming baselines in speed while guaranteeing safety.

BibTeX

@article{hong2026polymerge,
  title={PolyMerge: Compressing 3D Gaussian Splats with Polytope Coverings for Provably Safe Resource-Constrained Navigation},
  author={Hong, Jihoon and Chiu, Chih-Yuan and Fridovich-Keil, Sara and Chou, Glen},
  journal={IEEE Robotics and Automation Letters},
  year={2026},
  publisher={IEEE}
}