GraphWalker Indoor Localization Dataset: Raw Smartphone IMU Trajectories, Ground-Truth Waypoints, and Semantic Floor Plan Graphs
收藏官方服务:
资源简介:
This dataset is shared as part and in support of research in infrastructure-free indoor localization, inertial odometry, and map-constrained trajectory matching. The dataset comprises 152 real-world indoor walking trajectories (approximately 170 minutes and 9,079 meters total) collected across three floor plan sections of varying structural regularity. Raw inertial data were recorded using a Samsung Galaxy S20 5G smartphone carried close to the chest, with participants following natural indoor walking behavior including straight-line traversals, multi-turn sequences, and brief pauses. Ground-truth waypoints were manually annotated by an observer following close behind each participant without influencing their motion.
提供机构:
Zenodo创建时间:
2026-05-25



