This paper presents an online learning algorithm to construct from video sequences an image-based representation that is useful for recognition and tracking. For a class of object...
We propose a simple framework that utilizes online appearance models for 3D face and facial feature tracking with a deformable model. Adapting the geometrical parameters for each ...
The phenomenal growth of video on the web and the increasing sparseness of meta information associated with it forces us to look for signals from the video content for search/info...
Ming Zhao 0003, Jay Yagnik, Hartwig Adam, David Ba...
A video sequence of a head moving across a large pose angle contains much richer information than a single-view image, and hence has greater potential for identification purposes....
Shaun Canavan, Benjamin Johnson, Michael Reale, Yo...
This paper presents a method to recover the full-motion (3 rotations and 3 translations) of the head from an input video using a cylindrical head model. Given an initial reference...