Face recognition in the presence of pose changes remains a largely unsolved problem. Severe pose changes, resulting in dramatically different appearances, is one of the main difï¬...
The variations of pose lead to significant performance
decline in face recognition systems, which is a bottleneck
in face recognition. A key problem is how to measure the
simila...
Annan Li (Chinese Academy of Sciences), Shiguang S...
A novel face recognition approach is proposed, based on the use of compressed discriminative features and recurrent neural classifiers. Low-dimensional feature vectors are extract...
Most face recognition systems only work well under quite constrained environments. In particular, the illumination conditions, facial expressions and head pose must be tightly con...
In this paper, we propose a new face recognition system based on the ordinal correlation principle. First, we will explain the ordinal similarity measure for any two images and th...
Ronny Tjahyadi, Wanquan Liu, Senjian An, Svetha Ve...