Penza State University, a member of the world-class scientific and educational center "Engineering of the Future", is developing an algorithm for processing speech signals to detect a person's psychoemotional state. Alan Alimuradov, Ph. D., an employee of the university, spoke about these developments:
"Each speech signal is saturated with an emotional component, and it can be very important to identify this component in order to understand what state a person is currently in when pronouncing speech. We are trying to apply this in highly responsible management systems. For example, for military services and the police. The employee's speech is monitored and analyzed in the background, their emotional state is determined, and a decision is made whether they are ready for work or not.
If he is not ready and does not understand it yet, the system, having detected this, sends a signal to higher authorities in order to prevent some emergency situations, thus excluding the human factor."
In a video interview, Alan Alimuradov talks in detail about what speech sources were used to test the algorithm, what indicators are analyzed to determine a person's psychoemotional state, and the difficulties encountered by developers in the process.
"But still, each person is unique, and the system is advisory in nature, that is, the final decision still remains with the person, with protein, and not with artificial intelligence..." The video is available at the link: https://youtu.be/Qqp5licucwQ