MAGICIAN: Efficient Long-Term Planning with Imagined Gaussians for Active Mapping

2026-03-23

AI Summary

Existing active mapping methods often suffer from low exploration efficiency and incomplete reconstruction due to their reliance on myopic next-best-view selection. This work proposes a long-horizon planning framework for active mapping that, for the first time, integrates implicit scene representations with strong structural priors and long-term trajectory optimization. By leveraging a pre-trained occupancy network to construct an “Imaginary Gaussian” scene representation, the approach enables efficient volume rendering and real-time estimation of surface coverage gains. A tree-search algorithm is employed to globally optimize trajectories, while the scene representation and trajectory are jointly updated in a closed-loop manner. The method achieves state-of-the-art performance across diverse indoor and outdoor benchmarks and under varying action spaces, demonstrating that long-horizon planning is crucial for enhancing both the completeness and efficiency of active mapping.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Active mapping aims to determine how an agent should move to efficiently reconstruct an unknown environment. Most existing approaches rely on greedy next-best-view prediction, resulting in inefficient exploration and incomplete scene reconstruction. To address this limitation, we introduce MAGICIAN, a novel long-term planning framework that maximizes accumulated surface coverage gain through Imagined Gaussians, a scene representation derived from a pre-trained occupancy network with strong structural priors. This representation enables efficient computation of coverage gain for any novel viewpoint via fast volumetric rendering, allowing its integration into a tree-search algorithm for long-horizon planning. We update Imagined Gaussians and refine the planned trajectory in a closed-loop manner. Our method achieves state-of-the-art performance across indoor and outdoor benchmarks with varying action spaces, demonstrating the critical advantage of long-term planning in active mapping.

Problem

• active mapping
• long-term planning
• next-best-view
• scene reconstruction
• exploration efficiency

Innovation

• long-term planning
• Imagined Gaussians
• active mapping
• volumetric rendering
• tree search