Computational Laban Movement Analysis using probability calculus
Joerg Rett, Jorge M.M. Dias · 2007
This work presents a system which implements the concept of Laban Movement Analysis (LMA) using probability calculus and Bayesian theory. Our Human-Interaction-Database (HID) provides sequences of position data from several persons performing several movements in 3-D (magnetic tracker) and 2-D (vision). From the position data a set of low-level features is calculated. Probability calculus is used to relate these low-level features and the frame of reference associated to the variables of LMA. The Bayesian theory provides the concept for calculation, learning and classification. 1