OGM2PGBM

The OGM2PGBM TUM CMS LiDAR long-term indoor localization dataset comprises raw LiDAR scan data simulated on a model of the offices at the Technical University of Munich, along with their corresponding 3D and 2D reference maps. The datasets were meticulously created using the Gazebo simulation engine, ensuring complete coverage of the space, as described in this https://publications.cms.bgu.tum.de/2022_ECPPM_Vega.pdf" paper. The dataset consists of six sequences, simulated in three different environments. Each environment introduces varying levels of deviations between the reference map (which includes only permanent structures) and the simulated data. Three sequences correspond to each map, while the other three include additional dynamic simulated humans, increasing the localization challenge. This dataset is highly valuable for researchers and developers aiming to enhance algorithms for long-term localization in dynamic indoor environments.