Magnetic sensor based indoor positioning by multi-channel deep regression

Abstract

We address the indoor localization problem, where the goal is to predict and propose a deep learning architecture for PDR (Pedestrian Dead Reckoning) together with based on fusing predictions WiFi and smart phone sensor data. We build a reliable prediction of the user’s trajectory from the data collected by their smartphone, using inertial sensors such as accelerometer, gyroscope and magnetometer, as well as other environment and network sensors such as barometer and WiFi. Our objective is to obtain a reliable prediction of the trajectory of a user from the data collected by his/her smartphone, using inertial sensors such as the accelerometer and the gyroscope, as well as other type of sensors such as barometer and WiFi scanner. Our system is based on four main components: Our system implements a deep learning based pedestrian dead reckoning (deep PDR) model that provides a high-rate estimation of the relative position of the user.

Using Kalman Filter, we correct the PDR’s drift using WiFi. Then, an indoor location system based on WiFi fingerprinting that provides a prediction of the user’s absolute position each time a WiFi scan is received, which occurs approximately every 4 seconds. A fusion of the two above-mentioned predictions using a Kalman filter. The output of this component is a new estimation of the user’s global position in which the possible physical restrictions imposed by the environment are not taken into account.
Finally, we adjust Kalman Filter results with a map-free projection method that takes into account the physical constraints of the environment (corridors, doors, etc.) and projects the prediction on the possible walkable paths.
We test our pipeline on IPIN’19 Indoor Localization challenge dataset and demonstrate that it improves the winner’s results by 20% using the challenge evaluation protocol.