Real-time pedestrian tracking by visual attention and human knowledge learning

Abstract

In this paper, a novel model of pedestrian tracking by using object-based attention and human knowledge is presented. The selective units in the system are the objects and groupings which are space-driven as well as feature-driven. The factors of speed, motion direction and spatial location are used to cluster and form the groupings. Hierarchical selectivity of attention for objects in a grouping is implemented under the guide of human model knowledge with the help of a head detector. The motion cues are utilized to tackle the multi-person tracking through hierarchical selection of attention. The experimental results from outdoor environments are reported.