Face recognition via statistical codebook modeling and matching

Abstract

This paper presents a novel face recognition method to reduce the influence form varying face postures while maintaining high computational efficiency. Via building a codebook of an ASM face model and adopting the Mann-Whitney U-test matching method, the input face can be compared with the database much efficiently and effectively. Experimental results show that the proposed method has better recognition result and less computational complexity, and is competitive and practical for implementing into the smart equipments.