Toward inertial position tracking for head-mounted displays: a dataset and a deep learning approach evaluation

This work addresses the problem of correcting the drift error produced by inertial sensors for tracking the position of a Head-Mounted Display in virtual reality applications. Unlike state-of-the-art works, which use exteroceptive sensors to address the problem, this work introduces a novel approach to virtual reality based on deep learning and using data exclusively from proprioceptive sensors. The main contributions of this work are: (1) generating a database with readings taken from an inertial measurement unit located on a virtual reality headset while the users perform different activities and (2) testing deep neural networks based on bidirectional LSTM layers to predict trajectories of the user’s head in virtual environments. Additionally, the results show that, compared to previous approaches based only on inertial readings, the proposed approach improves the estimation of the coordinate of the virtual reality headset corresponding to the user’s height.