Walking Direction Estimation using Smartphone Sensors: a Deep Network-Based Framework

Smartphone-based inertial and magnetic sensors can be the basis for pedestrian navigation, whenever external positioning signals are limited or unavailable. Such navigation solutions are typically accomplished by a practice known as pedestrian dead reckoning, wherein step length and heading angle are estimated to form the horizontal trajectory of the user. One of the main challenges in these methods is the unknown misalignment between user’s forward axis and device’s frame, which imposes great difficulty in finding the user’s heading. In this paper, the problem of estimating walking direction is addressed by a learning-based approach. Specifically, a novel deep network architecture is designed for extracting the motion vector in the device coordinates, using accelerometer measurements. It consists of temporal convolutions and multi-scale attention layers, and involves rotation-invariant property that is analytically derived. The network model is integrated with a geometric calculation, employing gravity and geomagnetic directions, to convert the motion vector into heading angle relative to north. Extensive experiments of natural walking activity were conducted by a single pedestrian, with smartphones placed freely in various pockets. These were used for training the proposed model, in a user-specific approach. Finally, the entire framework was evaluated on unseen data, comparing the deep network against a mechanistic baseline method. Using the proposed network, with merely 5500 parameters, the resulting heading errors had a median value of 9.8°, which was lower by 2.5° compared to the baseline method.