Class Mean based Face Recognition

摘要

In order to better exploit the class mean to classify the test sample, we propose a novel class-mean-based classifier for face recognition. This classifier first exploits a weighted sum of all the class means to express the test sample. The weight is determined under the constraint that the weighted sum has the minimum derivation from the test sample. Then the proposed classifier classifies the test sample into the class whose weighted mean is the closest to this sample. It seems that the metric used in the proposed classifier has the capability to better represent the relationship between the sample and the class, so it can obtain a much lower error rate than the minimum distance classifier.