ActiveGS: Active Scene Reconstruction Using Gaussian Splatting

TL;DR

ActiveGS tackles active scene reconstruction from RGB-D data by fusing a high-fidelity 2D Gaussian splatting map with a coarse 3D voxel grid to guide exploration and avoid collision. The approach introduces a confidence model over Gaussian primitives and uses candidate viewpoints around regions of interest to balance exploration and surface refinement, achieving improved reconstruction quality and coverage. The authors extend the evaluation with additional metrics, conduct ablations on the confidence formulation, discuss optimization and memory considerations, and demonstrate real-world UAV applicability. Overall, ActiveGS offers a practical framework for efficient, high-quality scene mapping with explicit considerations for exploration strategy and map fusion, enabling robust operation in unknown environments.

Abstract

Robotics applications often rely on scene reconstructions to enable downstream tasks. In this work, we tackle the challenge of actively building an accurate map of an unknown scene using an RGB-D camera on a mobile platform. We propose a hybrid map representation that combines a Gaussian splatting map with a coarse voxel map, leveraging the strengths of both representations: the high-fidelity scene reconstruction capabilities of Gaussian splatting and the spatial modelling strengths of the voxel map. At the core of our framework is an effective confidence modelling technique for the Gaussian splatting map to identify under-reconstructed areas, while utilising spatial information from the voxel map to target unexplored areas and assist in collision-free path planning. By actively collecting scene information in under-reconstructed and unexplored areas for map updates, our approach achieves superior Gaussian splatting reconstruction results compared to state-of-the-art approaches. Additionally, we demonstrate the real-world applicability of our framework using an unmanned aerial vehicle.