"Semantic-Enhanced LiDAR-Inertial SLAM Dataset"

"This dataset provides multimodal sensor recordings collected using a mobile robot platform for evaluating LiDAR\u2013inertial SLAM algorithms in degenerate environments. The dataset includes LiDAR point clouds, IMU measurements, and RGB camera images recorded in indoor corridor scenes characterized by weak geometric structures and strong self-similarity. All data are provided in ROS bag format and are accompanied by calibration parameters for sensor alignment. The dataset contains multiple sequences captured under different trajectories, enabling the evaluation of localization accuracy and robustness in challenging environments where traditional geometric constraints are insufficient. This dataset is intended to support research in SLAM, sensor fusion, and degeneracy-aware localization methods, and can serve as a benchmark for developing robust perception and mapping algorithms in structurally ambiguous environments."

本数据集提供了基于移动机器人平台采集的多模态传感器记录数据，用于在退化环境中评估激光惯性同步定位与地图构建（LiDAR-inertial SLAM）算法。该数据集涵盖激光雷达点云、惯性测量单元（Inertial Measurement Unit，IMU）测量数据与RGB相机图像，采集自几何结构薄弱、自相似性极强的室内走廊场景。所有数据均以ROS包格式提供，并附带用于传感器配准的校准参数。该数据集包含多组不同运动轨迹下采集的序列，可用于评估传统几何约束不足的挑战性环境中的定位精度与鲁棒性。本数据集旨在支撑同步定位与地图构建（Simultaneous Localization and Mapping，SLAM）、传感器融合以及退化感知定位方法相关研究，并可作为在结构模糊环境中开发鲁棒感知与建图算法的基准测试集。