Motion segmentation for humanoid control planning
Matthew Field, David A. Stirling, Fazel Naghdy, Zengxi Stephen Pan · Research Online (University of Wollongong) · 2008
The discovery of major management be-haviours from human motion data and uncov-ering their underlying components is investi-gated. A range of methods for segmenting major shifts in multidimensional time series are compared in inducing plausible behaviours from motion data. These behaviours are con-sidered as supersets of motion primitives that define a repertoire of manoeuvres available to the human. The resulting multilayered sym-bolic model is used as a framework for hu-manoid imitation and control. It is hoped that with appropriate matching and scaling of de-grees of freedom, models can be tested by ex-tracting a trajectory for a simulation of the Nao soccer bot. 1