This paper presents a new method to detect and accurately extract the moving object from a video sequence taken by a hand-held camera. In order to extract the high quality moving ...
Active Appearance Models (AAMs) are generative parametric models that have been successfully used in the past to track faces in video. A variety of video applications are possible...
Humans are articulated objects composed of non-rigid parts. We are interested in detecting and tracking human motions over various periods of time. In this paper we describe a met...
The appearance of non-rigid objects detected and tracked in video streams is highly variable and therefore makes the identification of similar objects very complex. Furthermore, i...
Although kernel-based methods have been demonstrated effectively in solving single-target tracking problem, facing more complicated multi-target tracking task, most of them still ...