Modeling the mindfulness people's function based on the recognition of biometric parameters by artificial intelligence elements

The subject matter of this article is the processes of modeling the function of attentiveness of users of critical applications on the basis of recognition of biometric parameters by elements of artificial intelligence. The goal is the development and software implementation of mechanisms for monitoring the work of employees of responsible professions, which, based on the analysis of information from a webcam online, monitor the presence of the employee's focus on the active zone of the critical application and the absence of unauthorized persons near the computer. The tasks to be solved are as follows: to determine a list of factors, the presence of which must be constantly checked to control the focus of the employee's attention on the active zone of the critical application and the absence of unauthorized persons near the computer; to choose the optimal technology for reading and primary processing of information from webcams online, for further use in solving the task; to develop mechanisms for monitoring certain factors, the presence of which must be constantly checked to control the presence of the employee's focus on the active zone of the critical application and the absence of unauthorized persons near the computer; and to programmatically implement the developed mechanisms using the object-oriented programming language Python. The methods used were artificial neural networks, 3D facial modeling, and landmark mapping. The following results were obtained. A list of factors has been identified, the presence of which must be constantly checked to monitor the presence of employee's attention in the active zone of critical use and the absence of unauthorized persons near the computer. On the basis of the analysis of modern technologies for reading and primary processing of information from online webcams, technologies implemented in the MediaPipe library were selected for further use in solving the problem. Mechanisms have been developed for monitoring certain factors, the presence of which must be constantly checked to monitor the presence of employees in the active zone of critical use and the absence of unauthorized persons near the computer. The object-oriented programming language Python is software-implemented using the MediaPipe library, mechanisms are developed and, based on the results of the experiments, the expediency of its use for solving the problem is proved. Conclusions. The scientific novelty of the obtained results is as follows: we have formed a list of factors, the presence of which must be constantly checked to monitor the presence of employee's attention in the active zone of critical application, the absence of unauthorized persons near the computer, and improved facial recognition technologies, which allows us to obtain a solution to the problem of monitoring the attention of users of critical applications in non-ideal conditions.