RT-GuIDE: Real-Time Gaussian Splatting for Information-Driven Exploration
Yuezhan Tao, Dexter Ong, Varun Murali, Igor Spasojevic, Pratik Chaudhari, Vijay Kumar
AI summary
Problem
Autonomous robots struggle to build high-fidelity, real-time maps while actively exploring unknown environments, as existing learned representations are too slow for onboard use or require memory-heavy auxiliary structures.
Approach
The framework unifies mapping and planning using 3D Gaussian Splatting, computing a fast uncertainty-based information gain metric and using a hierarchical GPU-accelerated planner to guide the robot toward high-uncertainty regions while ensuring collision-free navigation.
Key results
- Unified real-time mapping and planning framework built entirely on Gaussian splatting without auxiliary volumetric maps
- Fast information gain heuristic enabling real-time onboard viewpoint selection
- 18x speedup in trajectory planning via GPU acceleration compared to CPU implementations
- Superior map quality in simulation and real-world tests (≥0.8 dB higher PSNR, >16% better geometric accuracy) enabling open-set semantic segmentation
Why it matters
It bridges the gap between high-fidelity computer vision reconstruction and real-time robotic navigation, enabling autonomous systems to build and utilize detailed digital twins for complex exploration tasks.
Abstract
We propose a framework for active mapping and exploration that leverages Gaussian splatting for constructing dense maps. Further, we develop a GPU-accelerated 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 viewpoint selection experiments that our method yields comparable Peak Signal-to-Noise Ratio (PSNR) and similar reconstruction error to state-of-the-art approaches, while being orders of magnitude faster to compute. In closed-loop physics- based simulation and real-world experiments, our algorithm achieves better map quality (at least 0.8dB higher PSNR and more than 16% higher geometric reconstruction accuracy) than maps constructed by a state-of-the-art method, enabling semantic segmentation using off-the-shelf open-set models.