Compact Keyframe-Optimized Multi-Agent Gaussian Splatting SLAM

2026-04-01

AI Summary

This work addresses the challenge of inefficient communication in multi-agent SLAM under bandwidth-constrained conditions, where dense 3D Gaussian splatting maps hinder real-time collaborative mapping. The authors propose a communication-efficient multi-agent RGB-D Gaussian splatting SLAM framework that significantly reduces redundant data transmission through keyframe optimization and a novel Gaussian compression mechanism, achieving 85–95% reduction in communication volume while preserving rendering fidelity. A key innovation is a centralized loop closure detection method that operates without initial pose estimates and supports two alignment modes—pure rendered depth and camera depth—to balance accuracy and communication overhead. Experimental results on both synthetic and real-world datasets demonstrate substantial improvements in the practicality of multi-agent Gaussian splatting SLAM.

Abstract

Efficient multi-agent 3D mapping is essential for robotic teams operating in unknown environments, but dense representations hinder real-time exchange over constrained communication links. In multi-agent Simultaneous Localization and Mapping (SLAM), systems typically rely on a centralized server to merge and optimize the local maps produced by individual agents. However, sharing these large map representations, particularly those generated by recent methods such as Gaussian Splatting, becomes a bottleneck in real-world scenarios with limited bandwidth. We present an improved multi-agent RGB-D Gaussian Splatting SLAM framework that reduces communication load while preserving map fidelity. First, we incorporate a compaction step into our SLAM system to remove redundant 3D Gaussians, without degrading the rendering quality. Second, our approach performs centralized loop closure computation without initial guess, operating in two modes: a pure rendered-depth mode that requires no data beyond the 3D Gaussians, and a camera-depth mode that includes lightweight depth images for improved registration accuracy and additional Gaussian pruning. Evaluation on both synthetic and real-world datasets shows up to 85-95\% reduction in transmitted data compared to state-of-the-art approaches in both modes, bringing 3D Gaussian multi-agent SLAM closer to practical deployment in real-world scenarios. Code: https://github.com/lemonci/coko-slam

Problem

• multi-agent SLAM
• Gaussian Splatting
• communication bottleneck
• 3D mapping
• bandwidth-constrained

Innovation

• Gaussian Splatting
• Multi-Agent SLAM
• Map Compaction
• Loop Closure
• Communication-Efficient Mapping