基于线性SVM子空间的正面人脸检测研究

Obverse Face Detection Based on the Subspace of Linear Support Vector Machine

摘要: 通过对二阶段的子空间方法的研究,提出了基于线性SVM子空间的正面人脸检测方法.首先构造线性SVM粗分类滤波器,然后在线性SVM粗分类滤波器分割的子空间内构造高斯核的非线性SVM分类器.检测时,为了加快速度引入了平均脸模板匹配进行粗筛选,然后依次通过线性SVM粗分类滤波器、非线性SVM分类器进行人脸检测.

Abstract: As an instance of the 2-stage Subspace Approach,a frontal method is proposed.A linear SVM classifier is trained as a filter to produce a subspace in which a non-linear SVM classifier with Gaussian kernel is trained for face detection.In detection procedure,a template matching is first used to coarse filtering for speeding up,then the face detection is given by the linear SVM filter and the non-linear SVM filter in turn.