RT-GuIDE: Real-Time Gaussian splatting for Information-Driven Exploration

2024-09-26

arXiv.org

AI Summary

To address real-time dense mapping and information-driven exploration for autonomous robots, this paper introduces Gaussian Splatting—previously unexplored in active mapping—into a unified framework that jointly optimizes map construction and motion planning. The method employs a photometric-geometric joint loss for online map refinement and integrates a parallelized information-gain-based motion planner to enable low-latency onboard navigation decisions and real-time situational awareness. Compared to conventional voxel- or grid-based information-gain approaches, our method achieves over an order-of-magnitude improvement in computational efficiency. Experiments demonstrate a 10% increase in PSNR and a 30% improvement in geometric reconstruction accuracy, while maintaining comparable information-gain performance in simulation. The core contribution lies in the first systematic application of Gaussian Splatting to active exploration, bridging high-fidelity dense mapping with efficient autonomous decision-making in a single, cohesive framework.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

We propose a framework for active mapping and exploration that leverages Gaussian splatting for constructing information-rich maps. Further, we develop a parallelized motion planning algorithm that can exploit the Gaussian map for real-time navigation. The Gaussian map constructed onboard the robot is optimized for both photometric and geometric quality while enabling real-time situational awareness for autonomy. We show through simulation experiments that our method is competitive with approaches that use alternate information gain metrics, while being orders of magnitude faster to compute. In real-world experiments, our algorithm achieves better map quality (10% higher Peak Signal-to-Noise Ratio (PSNR) and 30% higher geometric reconstruction accuracy) than Gaussian maps constructed by traditional exploration baselines. Experiment videos and more details can be found on our project page: https://tyuezhan.github.io/RT_GuIDE/

Problem

• Real-time dense mapping using Gaussian splatting for exploration
• GPU-accelerated motion planning for real-time navigation
• Optimizing photometric and geometric map quality for autonomy

Innovation

• Real-time Gaussian splatting for dense mapping
• GPU-accelerated motion planning for navigation
• Optimized photometric and geometric map quality