Grid-Based multi-scale PCA method for face recognition in the large face database

Authors:
Haiyang Zhang;Huadong Ma;Anlong Ming
Affiliations:
Beijing Key Lab of Intelligent Telecommunications Software and Multimedia, School of Computer Science & Technology, Beijing University of Posts and Telecommunications, Beijing, China;Beijing Key Lab of Intelligent Telecommunications Software and Multimedia, School of Computer Science & Technology, Beijing University of Posts and Telecommunications, Beijing, China;Beijing Key Lab of Intelligent Telecommunications Software and Multimedia, School of Computer Science & Technology, Beijing University of Posts and Telecommunications, Beijing, China
Venue:
APWeb'06 Proceedings of the 2006 international conference on Advanced Web and Network Technologies, and Applications
Year:
2006

Citing 3
Cited 0

mmGrid: Distributed Resource Management Infrastructure for Multimedia Applications

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
Unified Subspace Analysis for Face Recognition

ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
The Anatomy of the Grid: Enabling Scalable Virtual Organizations

International Journal of High Performance Computing Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we propose an efficient grid-based multi-scale PCA method in the large face database. This method divides the large face database into some small sub-face databases by maximizing the variance in the face sub-database and minimizing the variance between the face sub-databases, then it segments the recognition process into the local coarse profile recognition process and accurate detailed geometric sub-component analysis process, and assigns the local coarse profile recognition process to the nodes of the multimedia service grid to reduce the recognition time. Our experimental results show that with the increase of the face database, this method not only reduces the recognition time, but also remarkably increases the recognition precision, compared with other PCA methods.