Face recognition with support vector machine
Shaoyan Zhang, Hong Qiao · 2004
The application of support vector machines (SVMs) in face recognition is investigated in this paper. SVM is a classification algorithm developed by V. Vapnik and his team. Based on the underlying optimization and statistical learning theories, SVMs provide a new approach to the problem of pattern recognition. In this paper, both linear and nonlinear SVM training models are used in face recognition. Faces in different orientations are taken as training samples. Primary results show that nonlinear training machine is better than linear machine; the former one always has a much larger margin, which means that it has a much stronger ability in classification and recognition.