Modified Quadratic Discriminant Functions and the Application to Chinese Character Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Probabilistic Visual Learning for Object Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Regularized discriminant analysis for the small sample size problem in face recognition
Pattern Recognition Letters
Working Set Selection Using Second Order Information for Training Support Vector Machines
The Journal of Machine Learning Research
Journal of Cognitive Neuroscience
Kernel quadratic discriminant analysis for small sample size problem
Pattern Recognition
Discriminative learning quadratic discriminant function for handwriting recognition
IEEE Transactions on Neural Networks
High-speed face recognition based on discrete cosine transform and RBF neural networks
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
Small sample size (SSS) problem is usually a limit to the robustness of learning methods in face recognition. Especially in the quadratic discriminant functions (QDF), too many parameters need to be estimated and covariance matrix of a class is usually singular. In order to overcome the SSS problems, we proposed a novel approach called orthogonal quadratic discriminant functions (OQDF). The OQDF assumes probability distribution functions of each two classes of face images have a uniform shape. Then, three OQDF models are developed. The Laplacian smoothing transform (LST) and Fisher's linear discriminant (FLD) are employed to preprocess the face images for the OQDF classifier. Finally, we evaluate our proposed algorithms on two face databases, ORL and Yale.