occupancy mapping

Zero-shot semantic 3D occupancy reconstruction from unlabeled multi-sensor data remains challenging due to the absence of semantic supervision and effective joint geometric-semantic modeling. Method: This paper proposes a scene-aware geometry-semantic joint reconstruction framework. Its core components include: (1) a novel Gaussian lattice modeling paradigm that jointly perceives semantics and geometry; (2) a cumulative Gaussian voxel splatting algorithm enabling efficient, differentiable voxelization; and (3) integration of vision-language model priors, LiDAR-guided geometric constraints, and jointly parameterized Gaussian representations—entirely without semantic annotations. Results: Under zero-shot settings, our method achieves state-of-the-art performance, substantially outperforming existing self-supervised approaches while attaining semantic occupancy accuracy close to fully supervised baselines. It establishes a new paradigm for open-world 3D understanding.

This work addresses the excessive memory consumption of conventional occupancy grid maps when deployed at high resolutions or large scales. To mitigate this, the authors propose a boundary-based map representation that explicitly stores only boundary voxels—such as occupied and frontier voxels—while implicitly encoding interior free and exterior unknown regions via two-dimensional closed surfaces. An efficient mechanism for occupancy state querying and updating is devised, integrated within a global–local mapping framework to enable real-time construction. Additionally, specialized data structures are introduced to enhance operational efficiency. The proposed method substantially reduces memory footprint while supporting efficient map construction, updates, and queries under real-time sensor input. The implementation has been made publicly available.

Traditional 3D occupancy grid mapping faces significant challenges in unknown environments, including high memory consumption and substantial update latency, which hinder its applicability for autonomous robots requiring efficient and scalable mapping. This work proposes a boundary-based occupancy mapping framework that innovatively integrates truncated ray casting with a direct boundary update mechanism. By eliminating the need for auxiliary local voxel grids, the method avoids storing voxels across the entire space and bypasses exhaustive ray traversal. Experimental results on public datasets demonstrate that the proposed approach substantially outperforms existing baseline and boundary-aware methods, achieving comparable mapping accuracy while significantly reducing both memory usage and update time.

Existing LiDAR occupancy grid prediction methods predominantly rely on deterministic, grid-level optimization, often yielding physically implausible and scene-inconsistent artifacts that compromise safety-critical autonomous navigation. To address this, we propose the first latent-space disentangled generative occupancy prediction framework: it decouples representation learning from stochastic prediction, explicitly modeling uncertainty in the latent space while supporting multimodal conditional inputs—including RGB images and high-definition maps. Our approach employs a hybrid VAE-GAN architecture, enabling end-to-end training and zero-shot cross-platform transfer. Evaluated on NuScenes, Waymo Open Dataset, and a proprietary real-world vehicle dataset, our method achieves state-of-the-art performance, significantly improving prediction fidelity, physical plausibility, and scene consistency—key requirements for robust autonomous driving systems.

This work addresses real-time, map-free self-localization on planar graphs (e.g., indoor floorplans) without retraining or large-scale image databases. The proposed data-driven, lightweight method operates solely on a generic monocular depth estimator and the floorplan. Its core contributions are: (1) a novel ray-based observation model that fuses single- and multi-view depth predictions, eliminating reliance on image upright orientation; and (2) a temporal-aware recursive Bayesian filtering module for efficient observation fusion and state update. The approach runs in real time on consumer-grade hardware. Evaluated on standard benchmarks, it significantly outperforms state-of-the-art methods [20, 28] in localization accuracy and robustness—especially under viewpoint and lighting variations—while requiring minimal deployment overhead and no domain-specific training.

Existing high-resolution 3D occupancy prediction methods are hindered by substantial computational costs and memory bottlenecks, making it challenging to achieve both accuracy and real-time performance. This work proposes a progressive multi-scale Gaussian occupancy prediction framework that overcomes the memory limitations of dense voxel representations through a hierarchical Gaussian seeding mechanism, which constructs a coarse-to-fine structure of Gaussian primitives. By integrating hierarchical Gaussian modeling, multi-scale feature fusion, and a hybrid sparse-dense representation, our approach enables end-to-end trainable and highly efficient inference. To support this research, we introduce TJScenes, the first panoramic six-camera dataset with fine-grained 0.1-meter occupancy annotations. Evaluated on both Occ3D-nuScenes and TJScenes, our method achieves state-of-the-art geometry reconstruction accuracy while significantly reducing inference latency, establishing a new Pareto frontier in the trade-off between efficiency and precision.

This work addresses the limitation of existing 3D scene graph methods, which treat perception as a post-processing step on static datasets and decouple scene understanding from observation planning, thereby hindering long-term, incremental environment modeling for robots. To overcome this, the paper introduces an online semantic exploration framework that, for the first time, formulates semantic scene completeness as an active optimization objective. The approach employs an uncertainty-guided traversal strategy to dynamically balance semantic verification, geometric coverage, and motion cost. By fusing RGB-D observations with a prior 2D occupancy map, it incrementally constructs an uncertainty-aware 3D scene graph encoding open-vocabulary object label posteriors and structural relational edges, which in turn drives closed-loop path planning. The system autonomously revisits semantically ambiguous regions and explores unknown spaces, enabling continuous, human-intervention-free patrolling, updating, and reasoning.

This work addresses the challenge of achieving long-term consistent semantic occupancy mapping across interconnected indoor spaces. To this end, we propose GEM-Occ, a novel framework that introduces a Gaussian evidence memory–based mechanism for semantic occupancy mapping. Our approach converts local visual predictions into semantic Gaussian occupancy and free-space ray evidence, which are then integrated into a hierarchical persistent memory structure—comprising local buffers, room-level subgraphs, and building-scale graphs—via a visibility- and uncertainty-aware causal update strategy. The framework further enables efficient Gaussian-to-occupancy splatting for rapid querying. Evaluated on the HIOcc benchmark, GEM-Occ substantially outperforms existing methods in local accuracy, online stability, free-space reasoning, revisit consistency, and building-scale scalability.

Reconstructing complete 3D occupancy scenes from visual observations in the presence of occlusions and unseen regions remains a significant challenge. This work proposes a sparse Gaussian occupancy representation that, for the first time, effectively integrates surface-centric visual geometric priors into voxel-level occupancy prediction. The method establishes a unified framework that jointly models multi-view and temporal information, enabling coherent scene understanding across both space and time. Evaluated on multiple indoor and outdoor benchmarks, the approach achieves state-of-the-art performance while maintaining computational efficiency, demonstrating strong generalization capabilities and consistent temporal coherence during inference.

This work addresses the lack of a consistent geometric evaluation criterion for room nodes in existing hierarchical 3D scene graphs, which leads to structural inconsistencies. To resolve this, the authors propose a novel approach grounded in occupancy decomposition that tracks free-space regions and generates polygonal footprints, thereby introducing free space as the first geometric anchor for room nodes. This strategy unifies both the construction and evaluation of the room layer. By integrating occupancy decomposition, free-space tracking, polygon generation, and matching with Matterport3D room instances, the method significantly improves room recall across twelve scenes. Although precision experiences a slight decline, the results highlight wall alignment boundaries as a persistent challenge across diverse environments.