File(s) not publicly available

Reason: Data available in other repositories.

Learning to Track Human Motion using Inertial Sensor Data

IMU data (specific force and angular velocities) and ground truth positional data for training AI algorithms to learn inertial sensor-based positional tracking.


Note: Some OxIOD sequences require processing to ensure input and output data are of same length. Rotation vectors of the ground truth orientation are: q_x, q_y, q_z, q_w convention.


OxIOD sample rate is 100Hz whereas RoNIN data set is sampled at 200Hz. Both use the Vicon motion capture systems.


A Hampel Filter was used to clean the ground truth data.

Funding

Industrial CASE Account - Cranfield University 2018

Engineering and Physical Sciences Research Council

Find out more...

History

Authoriser (e.g. PI/supervisor)

d.maton@cranfield.ac.uk