We describe a system to learn an object template from a video stream, and localize and track the corresponding object in live video. The template is decomposed into a number of lo...
Copyright 2004 IEEE. Published in Conference on Computer Vision and Pattern Recognition (CVPR-2004), June 27 - July 2, 2004, Washington DC. Personal use of this material is permit...
We propose a fast approach to 3?D object detection and pose estimation that owes its robustness to a training phase during which the target object slowly moves with respect to the ...
While there have been great advances in quantification of the genotype of organisms, including full genomes for many species, the quantification of phenotype is at a comparatively...
Peter Andrews, Haibin Wang, Dan Valente, Jih&egrav...
Abstract--Modeling background and segmenting moving objects are significant techniques for video surveillance and other video processing applications. Most existing methods of mode...