Statistical Person Verification Using Behavioral Patterns from Complex Human Motion

Abstract

Abstract. We propose a person verification method based on behavioral patterns from complex human movements. Behavioral patterns are represented by anthropometric and kinematic features of human body motion acquired by a Kinect RGBD sensor. We focus on complex movements to demonstrate that independent and rhythmic movement of body parts carries a significant amount of behavioral information. We take a statistical approach by Gaussian mixture models to model the individual behavioral patterns. We demonstrate that subject-preferred movements are more robust against forgery attacks and variations over time than predetermined subject-independent movements. The obtained equal error rate was 15.7% when using subject-preferred movements and 27.3% when using a predefined sequence of movements.

Learning shape variations of motion trajectories for gait analysis

“…More recently, the release of new depth sensors, like Microsoft Kinect offered new opportunities to analyze human motion, by providing in real-time 3D skeleton data representing the human pose. These data have been successfully used to support fall detection [16], action recognition [8], [24], motion segmentation [7], [25] and person authentication [1], [11].…”