GIAINDOORLOC

GIAINDOORLOC数据集是由亚琛工业大学大地测量研究所创建，包含2049条自动标记的指纹数据，覆盖了5层楼的室内环境。数据集通过结合WLAN和IMU数据，利用ARCore的VI-SLAM模块进行动态收集。创建过程中，通过智能手机应用实时记录位置和传感器数据，后处理阶段将数据映射到目标坐标系统。该数据集适用于训练基于神经网络的室内定位模型，旨在解决大规模室内环境中的中等至粗粒度定位问题。The GIAINDOORLOC dataset was developed by the Institute of Geodesy at RWTH Aachen University. It contains 2049 automatically labeled fingerprint entries, covering indoor environments across five floors. The dataset was dynamically collected using the VI-SLAM module integrated into ARCore by combining WLAN and IMU sensor data. During its construction, location and sensor data were recorded in real time via a smartphone application, and the collected data was mapped to a target coordinate system during the post-processing phase. This dataset is suitable for training neural network-based indoor positioning models, aiming to address medium-to-coarse grained localization problems in large-scale indoor environments.

搜集汇总

背景与挑战

背景概述

GIAINDOORLOC是一个用于室内定位研究的自动标注数据集，包含WLAN和IMU数据，通过VI-SLAM2tag方法生成。该数据集支持指纹识别室内定位，并提供了原始数据、标注数据及评估材料，适用于相关算法开发和性能评估。