Topological principal component analysis for face encoding and recognition |
| |
Affiliation: | 1. Department of Psychology, University of Tartu, Näituse-2, Tartu 50 409, Estonia;2. Department of Psychological Sciences, University of Missouri, United States |
| |
Abstract: | Principal component analysis (PCA)-like methods make use of an estimation of the covariances between sample variables. This estimation does not take into account their topological relationships. This paper proposes how to use these relationships in order to estimate the covariances in a more robust way. The new method topological principal component analysis (TPCA) is tested using both face encoding and recognition experiments showing how the generalization capabilities of PCA are improved. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|