A unique modernized complex of hidden inspection, which allows to determine not only the presence of dangerous objects and prohibited substances, but also in the future to distinguish fake human emotions from real ones, was presented at the international forum "Army-2021".
The complex was originally intended only for detecting dangerous objects (for example, weapons or drugs) in the flow of people. For the first time, the development was presented a year ago at the Army-2020 forum. "During the year, we have significantly improved this complex. The upgraded version is able to recognize six basic human emotions-anger, fear, sadness, joy, disgust and surprise.
As a result, we can make a conclusion based on the data of the inspection complex and the emotion recognition system, whether a particular person is dangerous or not," Nikita Bobkov, head of the scientific and thematic center "T" of JSC "NPK "NIIDAR", told TASS, adding that this will increase the reliability of detecting intruders. In parallel, scientists were able to significantly reduce the weight and size characteristics of the complex, and by as much as 30 times. "The reception part of the complex used to be placed in a 19-inch cabinet, now it fits into the dimensions of four packs of cigarettes," the agency interlocutor noted.
According to Bobkov, a demonstration sample of the complex's functionality operating in the millimeter and terahertz wave ranges has been designed and presented at the forum. The complex uses a neural network that reliably determines prohibited objects, as well as basic human emotions.
NIIDAR specialists plan to carry out further modernization of the complex: to develop a unique technology for identifying a person by gait, other external signs and individual characteristics of a person to detect suspicious behavior. In addition, the neural network will have to learn how to detect microexpressions of the face. The technology that can determine whether a person sincerely expresses emotions or imitates them should appear in 2022.
Ivan Petrov
