Parameter Estimation from Motion Tracking Data

TL;DR: The Bayesian filtering technique provides an efficient way to obtain the distributional estimate of the unknown parameters of the users based on tracking data and is well-suited to identifying parameters of articulated models in the presence of noisy data.

Abstract: User tracking for gesture recognition, object manipulation and finger-based interaction within an immersive virtual environment represents challenging problems. The motion capture system is providing the data for the user’s motion recognition, but the uncertainty remains in obtaining the exact motion of the user due to the deformations, especially when the markers are attached to the clothes or to the skin. This paper address the question how can this uncertainty be solved, how can be obtained the geometrical parameters of the users based on tracking data. The tracking data obtained from markers cannot be independent and had to satisfy the physical constraint between the different body parts, represented by the joints of the human skeleton. The Bayesian filtering technique provides an efficient way to obtain the distributional estimate of the unknown parameters. The obtained algorithm is well-suited to identifying parameters of articulated models in the presence of noisy data.