Abstract:Practical video scene and face recognition systems are sometimes confronted with low-resolution (LR) images. The faces may be very small even if the video is clear, thus it is difficult to directly measure the similarity between the faces and the high-resolution (HR) training samples. Traditional super-resolution (SR) methods based face recognition usually have limited performance because the target of SR may not be consistent with that of classification, and time-consuming SR algorithms are not suitable for real-time applications. In this paper, a new feature extraction method called Coupled Kernel Embedding (CKE) is proposed for LR face recognition without any SR preprocessing. In this method, the final kernel matrix is constructed by concatenating two individual kernel matrices in the diagonal direction, and the (semi-)positively definite properties are preserved for optimization. CKE addresses the problem of comparing multi-modal data that are difficult for conventional methods in practice due to the lack of an efficient similarity measure. Particularly, different kernel types (e.g., linear, Gaussian, polynomial) can be integrated into an uniformed optimization objective, which cannot be achieved by simple linear methods. CKE solves this problem by minimizing the dissimilarities captured by their kernel Gram matrices in the low- and high-resolution spaces. In the implementation, the nonlinear objective function is minimized by a generalized eigenvalue decomposition. Experiments on benchmark and real databases show that our CKE method indeed improves the recognition performance.

Kernel linear regression for face recognition

Face recognition using laplacianfaces.

Learning a Locality Preserving Subspace for Visual Recognition.

Face recognition with contiguous occlusion using linear regression and level set method

Bilateral Structure Based Matrix Regression Classification for Face Recognition

Face recognition using a kernel fractional-step discriminant analysis algorithm

Center-based weighted kernel linear regression for image classification.

Kernel optimization-based discriminant analysis for face recognition

Adaptive Linear Discriminant Regression Classification for Face Recognition

Modified Kernel-Based Nonlinear Feature Extraction [face Recognition Example]

Kernel collaborative representation-based classifier for face recognition

An Efficient Reformative Kernel Discriminant Analysis for Face Recognition

Kernel Neighborhood Preserving Projections for Face Recognition

Fuzzy Linear Regression Discriminant Projection for Face Recognition

Face Recognition Using Nonlinear Partial Least Squares in Reproducing Kernel Hilbert Space

Kernelized dual regression incorporating local information for image set classification

Kernel Generalized Nonlinear Discriminant Analysis Algorithm For Pattern Recognition

Recognizing face or object from a single image: linear vs. kernel methods on 2d patterns

Coupled kernel embedding for low resolution face image recognition

Local median based linear regression classification for biometric recognition