Existing psychological theories connect movements of human eye pupils with its reactions to external changes: what we see, hear, feel. Analysing the look, we can compare human external reaction, i.e. what his behavior shows, and natural reaction, i.e. what he really feels.
Currently, practically each organization faces the necessity of estimation of psychoemotional human state. It is especially acute at schools and universities, where teachers work with children and teenagers, and at jobs connected with great risks or human lives.
For this aim companies use different methodologies of candidate diagnostics, for example, a polygraph, but a person does not feel comfortable, which adversely affects test results. Another way of estimation of psychoemotional human state is the analysis of his visual activity, based on the interaction of the movement of an eye’s pupil with its reaction to external changes.
Previously a lot of approaches were offered for detection of an eye at a videoimage, which used different characteristic peculiarities of the initial data. These methods can also be applied for the current task, but their calculations are too complicated for it, require large consumption of main storage and education of the recognizer, unjustified power enlargement and complexity of the hardware. Moreover, the results of the education are not always easily interpreted and usually can not be broadened to similar tasks without retraining.
Considering human physiological peculiarities, specialists of the Institute of Cyber Intelligence Systems and MEPhI Engineering Centre have developed a complex for recognition of vision activity and its application for estimation of psychophysiological human state. The suggested system allows raise the precision of recognition of the human look direction, saving simplicity of the used hardware and rational usage of the hardware power.
The complex consists of a videocamera, placed at glasses and working in near infrared range, and developed software. Glasses with a camera record all the movements of a human eye in real-time mode. Receive/transmit and processing devices get, transform and transfer data from cameras to research equipment, which conduct main stages of the pupil recognition and interpretate data of the recognition in accordance with the asked question.
The result is given in informative and clear report, which can be used for further analysis.
Currently, there is the first version of the device prototype, authors have conducted the correction of the software up to automated diagnostics level. The technology is going through additional testing, in which psychological theory is being proved.
The complex is user-friendly and shows high efficiency in the estimation of psychoemotional human state. It can be used for raising of accuracy of the research, connected with detection of human reactions in: interview for the job, analysis of the psychological climate in a group; psychological tests and research; constant monitoring of psychoemotional human state; marketing surveys.
The article about the development of MEPhI scientists has been published in AIP Conference Proceedings, which is in the lists of editorials, cited by Scopus and Web of Science.
Photo from homeplanzoom.us
