Evolutionary pursuit and its application to face recognition

This paper introduces Evolutionary Pursuit (EP) as a novel and adaptive representation method for image encoding and classification. In analogy to projection pursuit methods, EP seeks to learn an optimal basis for the dual purpose of data compression and pattern classification. The challenge for EP is to increase the generalization ability of the learning machine as a result of seeking the trade-off between minimizing the empirical risk encountered during training and narrowing the confidence interval for reducing the guaranteed risk during future testing on unseen images. Toward that end, EP implements strategies characteristic of genetic algorithms (GAs) for searching the space of possible solutions to determine the optimal basis. EP starts by projecting the original data into a lower dimensional whitened Principal Component Analysis (PCA) space. Directed but random rotations of the basis vectors in this space are then searched by GAs where evolution is driven by a fitness function defined in terms of performance accuracy (`empirical risk') and class separation (`confidence interval'). Accuracy indicates the extent to which learning has been successful so far, while separation gives an indication of the expected fitness on future trials. The feasibility of the new method has been successfully tested on face recognition where the large number of possible bases requires some type of greedy search algorithm. The particular face recognition task involves 1,107 FERET frontal face images corresponding to 369 subjects. To assess both accuracy and generalization capability, the data includes for each subject images acquired at different times or under different illumination conditions. The results reported show that EP improves on face recognition performance when compared against PCA (`Eigenfaces') and displays better generalization abilities than the Fisher linear discriminant (`Fisherfaces').