This article analyzes known data on the systemic functions of vision, such as discrimination and recognition of visual objects, visual seeking, assessment of the emotional content of scenes, and decision-making in the foveal and peripheral visual fields. Existing hypotheses for the possible mechanisms of functional phenomena occurring in the human peripheral field are discussed. A neurological informatic approach to solving problems of the interaction of foveal and peripheral vision based on inspection trajectories, areas of interest, and return gaze fixations are described. Computer experiments showed that the structure of an inspection trajectory model correlates with the number of return fixations of the “input window” of the model. This suggested that the probability of return fixations could be regarded as a quantitative criterion for identifying the type of attention (focal or spatial) and the moment of attention-switching.

An overview of known works in active vision area and our recent results on application of the foveal visual preprocessor to detect the head motion parameters are presented. In overview, the main directions of research and development in the field of artificial foveal active vision have been considered. It is justified that: (i) for a successful solution of complex problems in this area and creation of universal systems based on active foveal vision, it is necessary to develop new technologies and platforms for the experimental study of various aspects of active foveal vision in detail; (ii) at present, software implementations of the foveal principles of visual information processing already contribute to solving particular applied problems of computer vision. In computer simulation, detection of the initial moment of head motion was evaluated by means of foveal visual preprocessor. In this neural network, each pair of excitatory and inhibitory neurons has common center, different sizes of their receptive fields and time delay. To test network performance, synthetic video of facial image sequences from SYLAHP database monitoring the head motion were used. It was shown that the U-E amplitude and polarity qualitatively correspond to face motion amplitude and direction. The initial front of U-E changes corresponding to quick motion of head was equal to 12 ms in all cases (n = 46). Video of real face images with graduated turns was tested too to estimate quantitative relation between output function of excitatory neuron and turn degree. It was revealed that this relation is equal to 40 U-E/degree. Future steps of research in this direction have been shortly discussed.

We analyzed a problem of determination of an optimal combination of algorithms and parameters of the remote photoplethysmography method for heart rate estimation. To solve this problem we developed an evolutionary algorithm implementing an adaptive reinforcement learning. The optimal solution of the problem of the remote photoplethysmography was obtained with the aid of an original criterion combining the mean square error with the dutation of agent’s life cycle.

The differences between trajectories of eye movements during navigation in a 3D virtual environment projected on a screen, and during 2D image viewing are considered in this paper. Two types of the tasks had been used in navigation tests. In the first one subject was navigate in a 3D virtual environment by pressing buttons. In the second the subject watched the environment during imposed motion of the virtual space. Differences between proportions of the fixations and smooth pursuits have been found: in the first type (active control) proportion of smooth pursuits was significantly larger than in the second type (passive control) - 67% and 55%, correspondingly. On the other hand, saccade amplitude, velocity and duration of fixation were similar in navigation tests. Typical patterns linked with camera rotation were revealed. In particular they were different in active and passive navigation tests: in active type, these patterns preceded the rotation, and in second type followed it. Also the movement's properties (fixation duration, spatial distribution of fixation on the image, peak speed and saccade amplitude) were different. Differences of eye movement properties (fixation duration, spatial distribution of fixation on the image, peak speed and saccade amplitude) were found in comparison of navigation tests and 2 D image viewing tests.

Several applications of artificial foveal visual systems in image processing are considered. The basic attention is focused on description of the Behavioral Model of Vision (BMV) developed in A.B. Kogan Research Institute for Neurocybernetics. The procedures for most informative region detection and space-variant context image representation are presented. The results of the BMV testing while recognition of facial (n=400) and traffic sign (n=202) images show a high recognition rate (95% for facial images and 97% for traffic signs) and demonstrate invariance to the point of view, noise and brightness for both image types. In addition, while processing PET situation images, the BMV detects facial landmarks (eye corners and middle point of nose basement) with very high accuracy (1±0.64 pixels).

FOSFI (Foveal System for Face Identification), a hardware/software system for authorized access by face personal identification, is presented based on biologically plausible algorithms of image description and recognition developed earlier. These algorithms perform detection of the most informative image regions, spa- tially nonuniform representation of visual information, and context encoding of primary features. The testing of the system in actual practice demonstrates that the system does not permit illegal access in 100% of cases, while the probability of false denials of an authorized access is less than 8%.