giaIndoorLoc – Auto-labeled WLAN + IMU dataset generated via VI-SLAM2tag

This repository holds the data that belongs to the publication: M. Laska, T. Schulz, J. Grottke, C. Blut and J. Blankenbach, "VI-SLAM2tag: Low-Effort Labeled Dataset Collection for Fingerprinting-Based Indoor Localization", [arXiv:2207.02668] which is to appear at the 2022 IPIN conference. It is split into the following sub-parts:<br> - giaIndoorLoc_raw: Raw data recorded via the VI-SLAM2tag android app (https://github.com/laskama/VI-SLAM2tag_app)<br> - giaIndoorLoc: Annotated dataset (generated from giaIndoorLoc_raw)<br> - evaluation_data: Raw trajectory data that is used during evaluation of labeling accuracy of VI-SLAM2tag (Control-Point + Total Station (Tachymeter))<br> - model_evaluation: Model weights of fitted models used during baseline performance section (VII-B) of paper. Required for reproducing experiments with repo (https://github.com/laskama/mCELindoorLoc) For a detailed description, please refer to the given paper and the additional github repositories that host the implementations: - https://github.com/laskama/VI-SLAM2tag_post - https://github.com/laskama/VI-SLAM2tag_app - https://github.com/laskama/mCELindoorLoc