Sample-Relaxed Two-Dimensional Color Principal Component Analysis for Face Recognition and Image Reconstruction

TL;DR

The paper introduces SR-2DCPCA, a novel color face recognition and image reconstruction method that uses relaxation vectors to improve recognition accuracy and image quality by enlarging global variance.

Contribution

It proposes a sample-relaxed 2DCPCA approach using quaternion models and relaxation vectors, enhancing recognition performance over existing methods.

Findings

Higher recognition rate than state-of-the-art methods

Efficient in image reconstruction

Validates effectiveness on real face datasets

Abstract

A sample-relaxed two-dimensional color principal component analysis (SR-2DCPCA) approach is presented for face recognition and image reconstruction based on quaternion models. A relaxation vector is automatically generated according to the variances of training color face images with the same label. A sample-relaxed, low-dimensional covariance matrix is constructed based on all the training samples relaxed by a relaxation vector, and its eigenvectors corresponding to the $r$ largest eigenvalues are defined as the optimal projection. The SR-2DCPCA aims to enlarge the global variance rather than to maximize the variance of the projected training samples. The numerical results based on real face data sets validate that SR-2DCPCA has a higher recognition rate than state-of-the-art methods and is efficient in image reconstruction.