Reduction in Computational Complexity-A Face Recognition Case Study

Abstract: Real time application like face recognition consist of loop carried dependencies. It includes a lot of data computation. As a result it leads to computational complexity. Due to computational complexity, execution time of a real time application like face recognition increases. Execution time of face recognition application can be reduced by changing the entire architecture of the system or it could reduce by modifying the existing algorithm. Changing the architecture of the system is not possible but modifications in the existing algorithm can be done. The main objective of this work is to reduce the execution time in a real time application like face recognition, thereby reducing the computational complexity. For this purpose, the existing algorithms used for face recognition is modified.

Keywords: Eigen Values, Eigen vectors, Threshold, Euclidean distance, PCA, LDA, Computational Complexities, cannonical correlation, Factor Analysis, Covariance, Scaling, Execution time.