SGE: Semantically-Guided Exploration for Unstructured Environments via Image-Space Waypoint Sampling
RoboticsArtificial Intelligence
Summary
The authors developed a new method called Semantically-Guided Exploration (SGE) to help ground robots explore environments better. Instead of just looking at shapes and obstacles, their method considers what different parts of the scene are (like terrain type or objects) using image analysis. Robots pick exploration goals by combining these insights with 3D mapping and real-time planning, allowing smarter navigation and the ability to recover from problems. They tested SGE in simulations and real indoor and mining environments, showing it performs well across different robots and settings.
Authors
Christopher Tatsch, Yu Gu
Abstract
This work introduces Semantically-Guided Exploration (SGE), a modular exploration framework for ground vehicles that integrates pixel-level semantic segmentation into sampling-based waypoint selection and receding-horizon route optimization. Unlike conventional geometric exploration methods, SGE evaluates candidate exploration goals directly in the image space using a semantic-aware utility function that accounts for terrain traversability, obstacle proximity, objects of interest, and depth-based exploration reward. Sampled waypoints are projected into 3D and ordered through a real-time Traveling Salesman Problem (TSP) formulation, enabling receding-horizon goal selection. To address real-world navigation uncertainty, the framework introduces mechanisms, including temporary taboo regions to handle navigation failures and a graph-based relocation strategy for efficient backtracking across explored areas. We evaluate SGE in standardized simulation benchmarks against state-of-the-art exploration planners and demonstrate competitive performance in volumetric coverage, while enabling semantic task biasing that cannot be achieved by purely geometric methods. The framework is further validated through real-world experiments using multiple robotic platforms in indoor campus buildings and in limestone and coal mines. Results show consistent performance and adaptability across platforms and domains.