Supervised Locality Preserving Projection Based Face Recognition
Xuemei Chen · Jisuanji fangzhen · 2010
Face recognition is an important branch of computer vision and can be applied in lots of fields.In the past few years,face recognition has received more focuses.Dimensionality reduction is the key technique for face recognition.It not only can clear some unnecessary noise,also can reduce the complicated degree of computation effectively.The traditional Locality Preserving Projection algorithm only considers the local information of the example data and neglects the impact of the different class labels,which results in the overlap of the data points of different classes.In this paper,we introduce the class labels information based on LPP,and propose Supervised Locality Preserving Projections(SLPP).Through some face recognition experiments,we show the effectiveness and robustness of SLPP.