Jul 2026· 2026 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)· pp. 1-6· 0 citations· 14 references
Abstract
This paper presents a semantic mapping system that integrates Vision-Language Models (VLMs) with Simultaneous Localization and Mapping (SLAM) to enhance the understanding of indoor environments. While traditional SLAM systems primarily focus on geometric occupancy, they often lack the semantic context needed for high-level robotic tasks. The proposed framework bifurcates the mapping process into two synchronized subsystems: Geometric SLAM and VLM Recognition. The first system uses LiDAR and odometry data to perform real-time localization and mapping, providing a stable and accurate environment for semantic data integration. Simultaneously, the second system leverages a VLM with a ZED2 depth camera to identify furniture and perform precise object size estimation, mapping these semantic entities into the global coordinate frame. A data fusion layer then synthesizes these inputs into a comprehensive Semantic Map that includes both geometric structures and physical object dimensions. Experimental results demonstrate that our integrated approach achieves precise spatial anchoring of semantic entities and improves overall mapping stability, providing a robust foundation for robot environmental understanding in complex indoor scenarios.
Simultaneous Localization and Mapping (SLAM) is essential for autonomous systems navigating in human-centric environments, yet conventional systems fail when people and objects move through the scene. This paper introduces SAR-SLAM (Semantic-Aware Recognition SLAM), an RGB-D SLAM framework that robustly handles dynamic scenes containing moving people and objects using dual semantic geometric processing. First, we employ YOLOv8-based semantic segmentation to identify dynamic objects and generate initial detection masks. Second, we apply RANSAC-based Homography analysis to perform geometric motion verification, distinguishing truly moving objects from stationary ones by analyzing feature correspondence patterns. Third, an adaptive fusion mechanism combines both semantic and geometric evidence while incorporating temporal consistency and coverage constraints to maintain system stability. The system is implemented as a modular ROS2 package, enabling smooth integration with robotic systems and compatibility with existing navigation frameworks. SAR-SLAM reduces Absolute Trajectory Error by up to 96% over ORB-SLAM3 on the dynamic sequences of the TUM RGB-D benchmark, and remains competitive with state-of-the-art dynamic SLAM methods across a range of dynamic scenarios.
Basheer Al-Tawil, Magnus Jung, T. Hempel et al.· Robotics· 0 citations
Mainstream LiDAR simultaneous localization and mapping (SLAM) systems often face challenges such as cumulative localization drift and loop closure mismatches when operating in large-scale complex outdoor environments, making it difficult to maintain global map consistency. This paper aims to propose a LiDAR SLAM system that integrates ground constraints with semantic information to comprehensively enhance robustness and localization accuracy in complex scenarios.
First, to address pose estimation errors caused by complex terrain, a two-step ground segmentation strategy is proposed. This method obtains reliable ground parameters through coarse extraction and fine fitting, and introduces ground constraint factors into the backend pose graph optimization, effectively suppressing cumulative system drift. Second, addressing the issue of perceptual aliasing in traditional geometric loop closure detection within dynamic or structurally similar scenes, a coarse-to-fine two-stage loop closure detection method is proposed. The first stage uses Scan Context for rapid candidate frame retrieval, while the second stage integrates semantic topological features with geometric distributions for precise matching and verification, thereby eliminating false matches and calculating high-precision 6-DoF loop closure poses.
Experimental results on a quadruped robot platform and the KITTI public data set demonstrate that the proposed method significantly reduces trajectory errors while maintaining real-time performance, showing superior performance in handling slopes, dynamic environments and large-scale loop closure scenarios.
This work introduces a robust outdoor LiDAR SLAM framework that effectively combines ground constraints with a novel coarse-to-fine semantic loop closure detection method to resolve elevation drift and perceptual aliasing in complex environments.
Yongchao Fan, Tongfei Hu, Fan Zhao et al.· Industrial robot· 0 citations
Autonomous semantic indoor navigation requires robust low-level control and high-level understanding of objects and spatial context in cluttered and partially occluded environments. While deep reinforcement learning (DRL) methods such as twin delayed deep deterministic policy gradient (TD3) enable reactive obstacle avoidance, they typically struggle with long-horizon semantic navigation, where object-location memory and language-level reasoning are required. We present a lightweight hierarchical two-stage framework that, to the best of our knowledge, is introduced for the first time to integrate a locally deployed vision–language model (VLM), semantic object coordinate memory, and a TD3-based DRL controller for language-conditioned indoor navigation. In Stage 1, the robot performs semantic exploration using odometry, 2D LiDAR, and VLM-based object recognition to build a geometric map and store detected object categories with their estimated world coordinates in a structured javaScript object notation (JSON) semantic memory. In Stage 2, a natural language query is used to retrieve the target object coordinates from memory and pass them to a TD3 target point navigation policy, which performs mapless navigation using odometry and RealSense RGB-D perception. The proposed framework combines open-vocabulary VLM-based object coordinate estimation, LiDAR mapping, RGB-D perception, and language grounding within a unified semantic memory representation. Experiments in a ROS-integrated realistic simulation demonstrate consistent goal-reaching performance and improved navigation efficiency compared with an Artificial Potential Field baseline using the same VLM and a DRL + GPT-4o mini configuration. We also compare the proposed VLM-based recognition module with YOLO-World v2.6 and Grounding DINO, showing that the VLM-based approach provides more reliable semantic grounding and target-coordinate estimation in the tested indoor navigation scenarios, particularly for flexible natural language object queries.
Yernar Akhmetbek, Ayaulym Parmash, Temirlan Meiramkhanov et al.· Italian National Conference...· 0 citations
—Simultaneous Localization and Mapping (SLAM) is critical for Intelligent Transportation Systems (ITS), yet existing solutions struggle to balance accuracy, sensor versatility, and computational efficiency. Traditional monocular methods suffer from scale drift, while RGB-D systems with heavy semantic networks often exceed the computational budgets of onboard processors. To address these conflicting constraints, we propose Dual-Mode SPL-SLAM, a unified adaptive framework capable of adapting to different intelligent driving agents. The system intel-ligently switches between two operating paths: Mode I (Sensor-Semantic) leverages physical depth sensors and high-precision semantic segmentation for complex dynamic environments. To further improve accuracy, we use line-feature processing and a refined epipolar-constraint error for robust detection of known and unknown dynamic objects; Mode II (Neural-Clustering) employs a lightweight neural network (LiteMono) and K-means++ clustering within the object bounding box predicted by GPU-accelerated YOLOX to recover scale and filter dynamic objects using fast LK optical flow in monocular setups. Crucially, both modes converge into a shared Adaptive Point-Line Backend, which dynamically weights features based on environmental texture to ensure robustness. Extensive experiments on KITTI dataset demonstrate that our system achieves state-of-the-art accuracy and real-time performance across diverse sensor configurations.
Zhen Tian, Zhihao Lin, Dezong Zhao et al.· IEEE transactions on intelli...· 0 citations
Large-scale indoor mapping and positioning with vision sensors is fundamental to a wide range of applications, such as robotic navigation and augmented reality. However, the rapidly increasing number of detectable objects and the expanded spatial coverage jointly introduce matching ambiguity and high computational cost. Fine-grained object maps can improve accuracy but often accumulate redundant observations and slow down localization, whereas overly compressed scene representations may discard essential semantic and structural cues and degrade robustness. To balance accuracy and efficiency for indoor spatial sensing, we propose TS-MapLoc, a map-centric object-level localization framework based on cross-layer semantic co-mapping. It builds a lightweight topological–semantic map that integrates multi-scale information from the image layer and the object layer, reducing redundancy while preserving key structural constraints. On top of this map, a cognition-inspired progressive localization strategy performs coarse-to-fine inference via stage-wise filtering under cross-layer semantic consistency, effectively narrowing the search space and stabilizing matching. The proposed method supports efficient and accurate object-level localization for built-environment applications.
Cuiyun Fang, Fan Wang, Ye-Dong Jiang et al.· IEEE Signal Processing Lette...· 0 citations
Achieving robustness, accuracy, and efficiency simultaneously remains a central challenge in light detection and ranging (LiDAR) simultaneous localization and mapping (SLAM). While learning-based approaches deliver strong benchmark performance, they often require extensive training, substantial computational resources, and struggle to generalize to unseen or degenerate environments. Geometry-based methods are efficient and interpretable, yet their performance degrades in planar or repetitive scenes due to limitations of standard iterative closest point (ICP) formulations. We present HP2-SLAM, a minimalist yet robust LiDAR SLAM framework built around a neighborhood-size adaptive hybrid ICP. Our key insight is a planarity-aware adaptive threshold that dynamically classifies correspondences based on local geometric structure and density, thereby enabling a principled balance between point-to-plane and point-to-point residuals. This formulation stabilizes alignment in both structured and degenerate environments without feature engineering, learning modules, or dataset-specific tuning. Integrated into a complete SLAM pipeline with submap management, loop closure detection, and pose graph optimization, HP2-SLAM consistently outperforms strong geometry-based baselines across publicly available datasets while maintaining real-time performance on commodity hardware. Our results demonstrate that carefully designed geometric adaptation can achieve strong generalization and robustness without sacrificing simplicity or efficiency.