Specially trained neural networks analyze the emotional state, pulse, respiration and gestures of the subject.
In Russia, they created a technology for remote monitoring of human emotions. To do this, using a large selection of Internet videos, neural networks were trained to “pull out” the largest number of parameters of a person’s internal state from videos. As a result, the algorithms were able to determine and analyze emotions, pulse, respiratory rate and gestures of people in the video.
The technology is useful in interviews before hiring, monitoring the status of drivers and pilots, as well as solving business problems where it is important to understand the opinion of the client.