基于遗传算法的人脸精确定位

提出一种基于遗传算法的人脸图象精确定位方法,作为人脸检测与人脸识别之间的一个中间优化处理过程,以解决由于人脸检测技术的定位误差导致人脸识别错误的问题。精确定位问题转化为一个在离散空间中的优化搜索问题,以遗传算法作为优化搜索方法,利用其收敛速度较快、鲁棒性强和全局寻优等优点,改进了基于身份子空间统一模型的人脸识别系统的识别效果。实验结果表明,基于遗传算法的人脸图象精确定位方法能够有效地提高人脸识别系统的识别率。

Precise face location method based on genetic algorithms

An original precise face location method,based on Genetic Algorithms(GA),is proposed to compensate localization errors,which often leads to face recognition failures.The precise face location problem was transformed to an optimal search problem in a discrete space.And then,the proposed GA-based method was applied to both solute the optimal search problem and improve the performance of the face recognition system based on the Unified Model in Identity Subspace(UMIS) approach.Experimental results demonstrated that the precise face location method based on GA could significantly improve the performance of the face recognition system.