Using Multiple Masks to Improve End-to-End Face Recognition Performance

Authors:
Christopher A. Neylan;Andrea Salgian
Affiliations:
Department of Computer Science, The College of New Jersey, Ewing, NJ 08628;Department of Computer Science, The College of New Jersey, Ewing, NJ 08628
Venue:
ISVC '08 Proceedings of the 4th International Symposium on Advances in Visual Computing, Part II
Year:
2008

Citing 4
Cited 0

Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection

IEEE Transactions on Pattern Analysis and Machine Intelligence
Recognizing Imprecisely Localized, Partially Occluded, and Expression Variant Faces from a Single Sample per Class

IEEE Transactions on Pattern Analysis and Machine Intelligence
Eigenfaces for recognition

Journal of Cognitive Neuroscience
Multiple face detection at different resolutions for perceptual user interfaces

IbPRIA'05 Proceedings of the Second Iberian conference on Pattern Recognition and Image Analysis - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we propose a method to improve performance in end to end face recognition systems. Our system uses a combination of masks that extract different regions of the face, and performs recognition separately on each region. Individual mask results are then combined through weighted sums and Borda voting. We test the method in conjunction with the fisherfaces (LDA) algorithm, and we analyze performance in an end-to-end system where face recognition is preceded by face and then eye detection using the Viola Jones algorithm. We find that our method improves recognition results by almost 10% on both manual and automatically detected face and eye locations.