A purely bottom-up model of visual attention is proposed and compared to five state-of-the-art models. The role of the low-level visual features is examined in two contexts. Two da...
This paper proposes a visual tracking algorithm that combines motion and appearance in a statistical framework. It is assumed that image observations are generated simultaneously ...
Aswin C. Sankaranarayanan, Rama Chellappa, Qinfen ...
Visual tracking is a challenging problem, as an object may change its appearance due to pose variations, illumination changes, and occlusions. Many algorithms have been proposed t...
In this paper, we propose a novel method to localize (or track) a foreground object and segment the foreground object from the surrounding background with occlusions for a moving ...
Lambert's model is widely used in low level computer vision algorithms such as matching, tracking or optical flow computation for example. However, it is well known that thes...