The best of Kalman-filter-based frameworks reported in the literature for rigid object tracking work well only if the object motions are smooth (which allows for tight uncertainty ...
Most of visual tracking algorithms have been achieved by matching-based searching strategies or detection-based data association algorithms. In this paper, our objective is to ana...
This paper argues that tracking, object detection, and model-building are all similar activities. We describe a fully automatic system that builds 2D articulated models known as pi...
The Automated Multiple View Inspection (AMVI) has been recently developed for automated defect detection of manufactured objects. The approach detects defects by analysing image se...
Off-line trained class-specific object detectors are designed to detect any instance of the class in a given image or video sequence. In the context of object tracking, however, o...