Golovan' A. V., Shevtsova N. A., Podladchikova L. N., Markin S. N., and Shaposhnikov D. G.
A.B. Kogan Research Institute for Neurocybernetics,
Rostov State University, Rostov-on-Don, Russia
Algorithms for identifying the most informative regions of face images (facial landmarks) as a concentration of the points of interest are presented. They are based on the neural network method for extracting local image features and constructing particular and composite feature maps. The algorithms developed demonstrate effective detection of the most informative regions invariant to changes in image view and size during processing the Olivetti Research Laboratory image database (coefficient of selectivity is greater than 80%).