Perception of human gestures through observing body movements

Abstract

A new approach to modelling and classification of human gait is proposed. Body movements are obtained using a sensor suit that records inertial signals that are subsequently modelled on a humanoid frame with 23 degrees of freedom (DOF). Measured signals include position, velocity, acceleration, orientation, angular velocity and angular acceleration. Using a range of concurrent features extracted from the sensor signals, a system using induced symbolic classification models, such as decision trees or rule sets, has been used for classification of identity. It is anticipated that this approach will also enable the identification of a variety of gestures. The feasibility of generating the identified behaviours in a humanoid robot will be explored. The approach is described and the characteristics of the algorithm are presented. The results obtained so far are reported and conclusions drawn.

Symbolic Modelling of Dynamic Human Motions

“…Apart from this, motioning the induced symbolic patterns also provides a diagnostic ability guiding the often cyclic and interactive nature of applying machine learning in general. Previous other studies have validated this approach by combining together with unsupervised mixture modelling for gait recognition (Field et al, 2008) (Hesami et al, 2008). The premise of this proposed work is that all humans have, by the stage of adolescence (or maturity) developed various stylistic signatures or patterns of motion behaviour that can be typically (uniquely) associated with an individual.…”