Face R的ecognition Eigen-faces with 99 PCA coefficients.pptVIP

Face RecognitionEigen-faces with 99 PCA coefficients Project Presentation Ashish Tiwari MAS622J/1.126J Pattern Classification and Analysis Problem Statement Given 2000 Training data points 2000 Test data points faceDR and faceDS: Description files for both training and testing data points. faceR: Matlab data file containing 99 PCA coefficients for training images faceS: Matlab file containing 99 PCA coefficients for test images. Required to classify Objective The Objective is to implement, test and compare performance of the various ‘ Pattern Classification Algorithms ’ on different facial features and to come up with a quantitative analysis of effectiveness of Algorithms in different scenarios. To develop insight of a designer’s perspective. Summary of the Work Parzen Window K- nearest neighborhood Generalized linear discriminant Neural network Boosting (neural nets as component classifiers) I have implemented all 5 Algorithms for almost all feature classification excluding Facial Expressions. Bad Data Issues As it invariably happens with any real-world problem data base, this problem also contains few bad faces, missing faces and outliers. In the implementation here the faces with missing descriptors are taken out from the data base where as outliers and bad faces are kept as it is. Eigen-facesAn Introduction Developed in 1991 by M.Turk Based on PCA Relatively simple Fast Robust Eigenfaces PCA seeks directions that are efficient for representing the data PCA reduces the dimension of the data Speeds up the computational time Eigenfaces, the algorithm Original Images Eigenfaces, the algorithm The mean face can be computed as: Mean-Face Eigenfaces, the algorithm Then subtract it from the training faces Eigenfaces, the algorithm Now we build the matrix which is N2 by M The covariance matrix which is N2 by N2 Eigenfaces, the algorithm Find eigenvalues of the covariance matrix The matrix is very large The computational effort is very big We are