Abstract
We propose a deep learning based framework that learns data-driven temporal priors to perform 3D human pose estimation from six body worn Magnetic Inertial Measurement units sensors. Our work estimates 3D human pose with associated uncertainty from sparse body worn sensors. We derive and implement a 3D angle representation that eliminates yaw angle (or magnetometer dependence) and show that 3D human pose is still obtained from this reduced representation, but with enhanced uncertainty. We do not use kinematic acceleration as input and show that it improves the generalization to real sensor data from different subjects as well as accuracy. Our framework is based on Bi-directional recurrent autoencoder. A sliding window is used at inference time, instead of full sequence (offline mode). The major contribution of our research is that 3D human pose is predicted from sparse sensors with a well calibrated uncertainty which is correlated with ambiguity and actual errors. We have demonstrated our results on two real sensor datasets; DIP-IMU and Total capture and have come up with state-of-art accuracy. Our work confirms that the main limitation of sparse sensor based 3D human pose prediction is the lack of temporal priors. Therefore fine-tuning on a small synthetic training set of target domain, improves the accuracy.
Highlights
Estimation of 3D human pose is an important goal in computer vision, augmented and virtual reality, robotics and human motion capture
Apart from uncertainty estimation, most important aspect of our work is that we develop a robust model and show that even a reduced orientation estimation from a sparse set of body worn magnetic-inertial measurement unit (MIMU) is ‘sufficient’ to estimate 3D human pose with enhanced uncertainty
Once we evaluated the results on individual sequences in test data, we found that mean per joint angle error (MPJAE) is high for certain sequences as shown in Fig. 4 and it is uncorrelated with the length of sequences
Summary
Estimation of 3D human pose is an important goal in computer vision, augmented and virtual reality, robotics and human motion capture. Inertial and magnetic sensors have become quite common with advent of low cost Microelectromechanical systems (MEMS) in recent years. This technology is called inertial motion capture (i-Mocap). Compared with camera based 3D pose estimation, body worn inertial motion capture is robust to occlusion and . Each sensor node comprises of magnetic-inertial measurement unit (MIMU), often called 9-axis IMU. It employs sensor fusion of rate gyro, accelerometer and magnetometer to obtain an orientation estimate and linear acceleration in a global frame. Many existing kinematic or inverse kinematic based i-Mocap frameworks, uses
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.