Y. Li, S. Gong, and H. Liddell, "Exploiting the Dynamics of Faces in Spatial-temporal Context", In Proc. International Conference on Control, Automation, Robotics and Vision, Singapore, December 2000.  Home/Search   Document Details and Download   Summary   Related Articles

....from sequences of rotating head is largely unresolved. Therefore, we had analyzed a method of Yongmin and Gong and simulated the effect of variation of some parameters to recognition ratio (using a program written in MATLAB) as well as have proposed improvement of basic matching criteria from [8]. 2. The eigenface method Consider the set of all possible images, those representing a face, make up only a small fraction of it. We decided to represent images as very long vectors, instead of the matrix representation. This makes up entire image space where each face is a point. Since faces ....

....a trajectory by projecting the face patterns into PCA space. On the other side, using the known prototype patterns, it is easy to construct trajectory for all known faces. Therefore, the recognition problem can be solved by matching the novel trajectory to a set of prototype trajectories. In [8] it is defined distance between the novel trajectory and a model trajectory m at each frame t as: d m = S w i d mi , i = 1, t (8) where d mi is the distance in feature space between model point and object point at frame i, and w is the weight on this distance. To achieve more robust and ....

Y. Li, S. Gong, and H. Liddell, "Exploiting the Dynamics of Faces in Spatial-temporal Context", In Proc. International Conference on Control, Automation, Robotics and Vision, Singapore, December 2000.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC