静态人脸识别的一种协同学方法

A Synergetic Algorithm for Static Face Recognition

本文针对静态人脸识别设计了一种简单易行的协同算法,它不需要面部特征的准确配准,当系统中的模式有增减时,算法可以快速获得新的原型向量和伴随向量。此外,对注意参数的两种情况的设置使算法的识别效果较单一平衡注意参数更好。仿真实验证明,该算法具有较好的识别效果。

A practicable algorithm for static face recognition is designed in this paper. Unlike the conventional methods, this algorithm doesn＇t depend on the accurate location of face features. While new patterns are added into the system, the algorithm can form the new original vectors and their adjoint vectors rapidly. In the simulation, two choices for attention parameters got good results.