Use of cluster validity in designing adaptive gabor wavelet based face recognition

Authors:
Eun Sung Jung;Phill Kyu Rhee
Affiliations:
Dept. of Computer Science & Engineering Inha University, Yong-Hyun Dong , Incheon, South Korea;Dept. of Computer Science & Engineering Inha University, Yong-Hyun Dong , Incheon, South Korea
Venue:
KES'06 Proceedings of the 10th international conference on Knowledge-Based Intelligent Information and Engineering Systems - Volume Part I
Year:
2006

Citing 7
Cited 0

Genetic Algorithms in Search, Optimization and Machine Learning

Genetic Algorithms in Search, Optimization and Machine Learning
Cluster validity methods: part I

ACM SIGMOD Record
Towards Automatic Face Identification Robust to Ageing Variation

FG '00 Proceedings of the Fourth IEEE International Conference on Automatic Face and Gesture Recognition 2000
Two-stage approach for pose invariant face recognition

ICASSP '00 Proceedings of the Acoustics, Speech, and Signal Processing, 2000. on IEEE International Conference - Volume 04
"Corefaces"- robust shift invariant PCA based correlation filter for illumination tolerant face recognition

CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
Illuminating light field: image-based face recognition across illuminations and poses

FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition
Some new indexes of cluster validity

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Face images in various situations due to facial expression, view point, illumination conditions, noise, etc. make identification process difficult. In this paper, the situation information of face images, what we call image context, is used to improve performance of a face recognition system. The proposed system partitions face images into several image contexts (groups) based on cluster validity, and takes adaptation to individual partitioned groups. In Gabor wavelet based face recognition, we apply weights to individual elements of facial feature, and those weights are trained by Genetic algorithm. We tried to use several unsupervised learning methods, clustering algorithms here, to partition face images into proper image contexts. There exists no formal way to decide the suitability of clustering algorithms for aiming at high recognition rate. We discuss about the process of cluster evaluation using the proposed cluster validity measure in designing adaptive face recognition. We achieved encouraging results though extensive experiments.