Ալեկսեյ Կուրին 018

___________________________________________________________________________

FORMATION OF SUBJECTIVE PORTRAIT BY MEANS OF NEURAL NETWORK BY VERBAL DESCRIPTION OF HUMAN APPEARANCE

 

Kurin Aleksey

Deputy Head Department of Criminalistics

of the educational and scientific complex

by preliminary investigation in the internal affairs bodies,

Volgograd Academy of the Ministry of the Interior of Russia,

Candidate of Technical Sciences, Associate Professor

__________________________________________________

 

Abstract: According to official statistics, the reason for not solving crimes in 98% of cases is the non-identification of the person who can be charged. Consequently, there is a practical need to search for new directions for obtaining orientation and search information that contributes to the identification of a person suspected of committing a crime. The success of the disclosure and investigation of crimes is determined not only by the presence, completeness and accuracy of information about the appearance of a person, but also by the efficiency of its receipt. In recent decades, the Russian security forces have adopted a large range of computer programs designed to compile a graphical model of a person's appearance. In the course of the study, a review of software and hardware systems designed to visualize information about the appearance of a person, used in the practical activities of the internal affairs bodies, was carried out, which made it possible to identify a combination of their advantages and disadvantages. The result of the study was a conclusion on the advisability of including in the list of information resources used to visualize a verbal description of the appearance of the wanted person, software products based on trained neural networks. The proposed approach will eliminate contradictions in the use of various terminology, as well as reduce the time spent on compiling a subjective portrait by using the capabilities of artificial intelligence systems.

Key words: verbal portrait, subjective portrait, orienting information, search information, accounting for subjective portraits.

 

PDF