融合AdaBoost和启发式特征搜索的人脸性别分类

Syncretize AdaBoost Learning and Heuristic Search to Select Features for Gender Classification of Frontal Facial Images

提出一种基于AdaBoost的人脸性别分类方法,从一张低分辨率灰度人脸图像中辨认出一个人的性别。将启发式搜索算法融于AdaBoost算法框架中,从而发现新的可用于更好分类的特征。利用该方法进行人脸性别分类方面的实验,当使用少于500个像素比较时,正确识别率达到了93%以上,这与迄今已公布的最佳的分类器支持向量机(SVM)的正确识别率相当,但速度却快得多。

This paper presents a method based on AdaBoost to identify the sex of a person from a low resolution grayscale picture of their frontal facial images. A heuristic search algorithm is used within the AdaBoost framework to find new features providing better classifiers, The experiment~ result of gender classification with the method presented in this paper indicate that the method is extremely last and achieves over 93% accuracy with less than 500 pixel comparisons operations, these match the accuracies of the SVM-based classifiers which the best classifiers published to date,