Human Face Recognition with Support Vector Machines: Algorithms and Applications

The computer vision problem of face recognition has over the years become a common high-requirement benchmark for machine learning methods. In the last decade, highly efficient face recognition systems have been developed that extensively use the nature of the image domain to achieve accurate real-time performance. The effectiveness of such systems are possible only with the progress of machine learning algorithms. Support vector machine learning is a relatively recent method that offers a good generalization performance in classification problems like face recognition. An algorithm based on Gabor texture information with SVM classifier is demonstrated in this book. The estimated model parameters serve as texture representation and experiments were performed on Yale, ORL and FERET databases to validate the feasibility of the method. The results showed that both Gabor magnitude and Gabor phase based texture representation technique with SVM classifier significantly outperformed the widely used Gabor energy based systems and other existing subspace methods. In addition, the feature level fusion of these two kinds of texture representations performs better than when used individually.