基于ICA/SVM的混合学习方法及在人脸检测中的应用

从基于样本学习的方法框架出发,该文提出了一种基于独立元分析和支持向量机的混合的学习方案,并用于人脸检测中。该方法通过独立元分析方法进行特征提取,然后采用SVM进行分类。该文做了大量的实验,以确定如独立元个数等参数问题对该分类器的影响,并与单独的支持矢量机方法、其它的人脸检测方法进行了比较分析。实验结果表明,该文的方法具有较好的检测效果,是一种很有效的方法。

Hybrid Learning Scheme Based on ICA/SVM and its Application in Face Detection

Starting from the frame of method based on learning sample,a hybrid learning scheme based on independent component analysis and support vector machine is proposed and used in face detection.In this method,independent component analysis is introduced into feature extraction,and support vector machine is used to classification of ICA feature.Authors have done a lot of experiments to get the influence of the number of independent component parameters on the classifier.And authors compare it with support vector machine,and other face detection methods by doing some experiments.The experiment results show that this approach can achieve good detection effect.It is proved to be a very effective method.