ARGMode — Activity Recognition using Graphical Models

This paper presents a new framework for tracking and rec- ognizing complex multi-agent activities using probabilistic tracking coupled with graphical models for recognition. We employ statistical feature based particle filter to robustly track multiple objects in cluttered environments. Both color and shape characteristics are used to differentiate and track different objects so that low level visual information can be reliably extracted for recognition of complex activities. Such extracted spatio-temporal features are then used to build temporal graphical models for characterization of these activities. We demonstrate through examples in dif- ferent scenarios, the generalizability and robustness of our framework.