ABSTRACT
This chapter reviews most commonly used visual features for describing human objects, which include Haar-like feature, scale-invariant feature features, and Histogram of orientfed gradient features. It examines a hierarchical Gaussian process latent variable model for human motion change detection. The chapter explores a Gaussian process dynamical model based on a unique social network feature set for modeling small human group behavior. Human motion modeling and motion change detection are important tasks in intelligent surveillance systems, and they have attracted significant amount of attentions. The major challenges in these tasks include complex scenes with a large number of targets and confusors, and complex motion behaviors of different human objects. The chapter demonstrates a multimedia content analysis system focusing on the group human action recognition, which includes feature extraction, human detection and tracking, and group human behavior classification.
