Virtual inertial measurements for motion inference in wireless health

Human motion monitoring with inertial sensors plays an important role in many medical applications. But, acquiring robust inference of human motion trajectory via low-cost noisy inertial sensors remains challenging. Sensor noise and drift, sensor placement errors and variation of activity over the population all lead to the necessity of a large amount of data collection. In this paper, a new virtual inertial measurement platform is proposed. With this platform, we are able to convert 3D camera data into noiseless and error-free inertial sensor measurements, thus permitting a large reduction in the quantity of expensive ground truth that would otherwise need to be collected.