Feature Patch Illumination Spaces and Karcher Compression for Face Recognition via Grassmannians  

作　　者：Jen-Mei Chang Chris Peterson Michael Kirby 

机构地区：[1]Department of Mathematics, Colorado State University, Fort Collins, USA [2]Departmentof Mathematics and Statistics, California State University, Long Beach, USA

出　　处：《Advances in Pure Mathematics》2012年第4期226-242,共17页理论数学进展（英文）

摘　　要：Recent work has established that digital images of a human face, when collected with a fixed pose but under a variety of illumination conditions, possess discriminatory information that can be used in classification. In this paper we perform classification on Grassmannians to demonstrate that sufficient discriminatory information persists in feature patch (e.g., nose or eye patch) illumination spaces. We further employ the use of Karcher mean on the Grassmannians to demonstrate that this compressed representation can accelerate computations with relatively minor sacrifice on performance. The combination of these two ideas introduces a novel perspective in performing face recognition.Recent work has established that digital images of a human face, when collected with a fixed pose but under a variety of illumination conditions, possess discriminatory information that can be used in classification. In this paper we perform classification on Grassmannians to demonstrate that sufficient discriminatory information persists in feature patch (e.g., nose or eye patch) illumination spaces. We further employ the use of Karcher mean on the Grassmannians to demonstrate that this compressed representation can accelerate computations with relatively minor sacrifice on performance. The combination of these two ideas introduces a novel perspective in performing face recognition.

关 键 词：GRASSMANNIANS Karcher Mean Face Recognition ILLUMINATION SPACES Compressions FEATURE PATCHES Principal ANGLES 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]