Clustering face images

In this paper a multi-step algorithm for clustering face images is proposed. This algorithm is designed to split a collection of images into groups of similar images. The algorithm is based on clustering the proximity measures between brightness-based segmented images. As proximity measures, the Euclidean distance and the Kullback-Leibler distance were used. Brightness-based image segmentation and clustering respective proximity measures were carried out with the help of a software model of a recurrent neural network. Results of experimental studies of the proposed approach are presented.