Home // SIGNAL 2022, The Seventh International Conference on Advances in Signal, Image and Video Processing // View article

Cleaning Outdoor Activity Logs using Deep Learning

Authors:
Davide Sbetti
Sergio Tessaris

Keywords: Data Cleaning, GPS Traces, Trajectory repairing, recurrent neural networks

Abstract:
Nowadays, position recording personal tracking devices are ubiquitous and used by both athletes and outdoor enthusiasts to track and analyse their activities. These devices rely on Global Navigation Satellite Systems to obtain the position in real time. Although the nominal precisions of the different GNSS are high enough for analysis, there are several environmental factors that affect the precision of such devices. Most of the commercial services providing analysis of outdoor activities use techniques to ``clean'' the user-uploaded data (tracklogs). Most of these techniques require and exploit the huge amount of data that they collect and analyse, but the resulting logs still manifest outliers and recording errors. In this paper, we present a deep learning based technique to identify part of tracklogs that might be influenced by recording errors, in such a way that can be corrected using standard techniques. Our approach does not require geographical or crowdsourced data, and can be also used on low powered devices.

Pages: 1 to 6

Copyright: Copyright (c) IARIA, 2022

Publication date: May 22, 2022

Published in: conference

ISSN: 2519-8432

ISBN: 978-1-61208-970-6

Location: Venice, Italy

Dates: from May 22, 2022 to May 26, 2022