A video surveillance system has been launched in Russian stores, which recognizes the thief before they commit theft

The Vidar company, together with Petrozavodsk State University, in early December 2024 presented a neural network video surveillance system capable of preventing theft in stores by analyzing the behavior of customers. The first tests showed an 80% reduction in losses.

The technology uses machine vision and artificial intelligence to analyze the behavior of store visitors in real time. The system is being tested at the Sigma at the House store of the Lotus trading holding in Petrozavodsk, and in February 2025 it is planned to be implemented in one of the federal retail chains.

𝓲

karelia.news

The development is based on the technology of detailed recognition of movements and position of a person's body with the construction of its "skeleton," which allows assessing psychophysiological deviations from the norm. The system analyzes ethological models and affective states of buyers.

Traditional video surveillance systems with analog cameras are often unable to detect suspicious behavior before the theft is committed. The new technology allows you to prevent offenses at the stage of preparation, and not fix them after the fact.

According to the publication, the Vidar company is the developer of a Russian neural network complex based on machine vision for retail. The use of domestic technologies makes it possible to ensure the independence of retail chains from foreign solutions in the field of security and loss prevention.

The innovative customer behavior analysis system demonstrates high efficiency in test stores and is preparing for large-scale implementation in the largest retail chains and shopping centers in Russia. Technology may become a future standard in the field of theft prevention in trade.^[1]

Notes