Visual learning and recognition of a probabilistic spatio-temporal model of cyclic human locomotion

We present a novel representation of cyclic human locomotion based on a set of spatio-temporal curves of tracked points on the surface of a person. We start by extracting a set of continuous, phase aligned spatio-temporal curves from trajectories of random points tracked over several cycles of locomotion in a monocular video sequence. We analyze a PCA representation of a set of cyclic curves, pointing out properties of the representation which can be used for spatio-temporal alignment in tracking and recognition tasks. We model the curve distribution density by a mixture of Gaussians using expectation-maximization algorithm. For recognition, we use maximum a posteriori estimate combined with linear data adaptation. We tested the algorithms on CMU MoBo database with favourable results for the recognition of people "by walking "from monocular video sequences captured from the side view.