Human Motion Change Detection by Hierarchical Gaussian Process Dynamical Model with Particle Filter

Human motion change detection is a challenging task for a surveillance sensor system. Major challenges include complex scenes with a large amount of targets and confusors, and complex motion behaviors of different human objects. Human motion change detection and understanding have been intensively studied over the past decades. In this paper, we present a Hierarchical Gaussian Process Dynamical Model (HGPDM) integrated with particle filter tracker for human motion change detection. Firstly, the high dimensional human motion trajectory training data is projected to the low dimensional latent space with a two-layer hierarchy. The latent space at the leaf node in bottom layer represents a typical human motion trajectory, while the root node in the upper layer controls the interaction and switching among leaf nodes. The trained HGPDM will then be used to classify test object trajectories which are captured by the particle filter tracker. If the motion trajectory is different from the motion in the previous frame, the root node will transfer the motion trajectory to the corresponding leaf node. In addition, HGPDM can be used to predict the next motion state, and provide Gaussian process dynamical samples for the particle filter framework. The experiment results indicate that our framework can accurately track and detect the human motion changes despite of complex motion and occlusion. In addition, the sampling in the hierarchical latent space has greatly improved the efficiency of the particle filter framework.