Our latest paper:
“Can I Trust This Location Estimate? Reproducibly Benchmarking the Methods of Dynamic Accuracy Estimation of Localization”.
has been published in the Journal Sensors by MDPI.
This work reviews the relevant literature related to the accuracy estimation of indoor positioning systems, presenting in a consistent terminology commonalities and differences of existing works and discussing baselines and evaluation metrics. Moreover, it presents a benchmarking framework, evaluating multiple relevant methods using open data, open code, and a rich set of relevant evaluation metrics. Our motive has been to establish the state of the art of this relatively overlooked but important feature of indoor positioning systems, in a reproducible manner.
The paper as well as the code implementation and the datasets used are openly available:
Paper: https://www.mdpi.com/1424-8220/22/3/1088
Code + Data: https://zenodo.org/record/5589651#.YfpQXS1Q1QJ
This work was co-funded by the Swiss National Science Foundation (SNSF) through the project Eratosthenes, grant number 195964, and by HES-SO, through the program «Mesures de soutien aux carriéres ralenties en raison de la crise du Covid».