Kaspersky Neural Networks

2021: Announcement of a recognition system for people, animals, buildings and cars from a drone

At the end of July 2021, Kaspersky Lab introduced a recognition system for people, animals, buildings and cars from unmanned aerial vehicles. The decision was called Kaspersky Neural Networks. According to the developers, it is intended mainly for the following purposes:

• search for missing people, animals and objects (e.g. flood-washed houses) in hard-to-reach areas;
• Assessment of damage in emergencies using aerial imagery;
• Analysis of data required for geodetic, cartographic, cadastral and traffic management activities.

Kaspersky Lab introduced a recognition system for people, animals, buildings and cars with UAVs

It is noted that Kaspersky Neural Networks provides real-time analysis of data from the board of the drone using neural networks. Kaspersky Lab cited one example of using a new product: during one of the Kaspersky Neural Networks tests, thousands of high-resolution images were processed in a few minutes on a laptop under conditions of working on rough terrain and limited computing power of the device.

According to the creators of the system, it will be relevant, first of all, for the needs of the Ministry of Emergencies and search and rescue teams. In large industrial enterprises, such as data centers, the solution can be used as a traffic control tool.

Experts interviewed by Vedomosti note that machine vision is a promising direction and the development of Kaspersky Lab is a strategically correct step. According to them, despite the lack of mass introduction of computer vision, there are requests and they will only grow.

According to a representative of Aeromax, drones with such a system can be estimated at 3.5-6 million rubles, exclusive for a certain company - 100 million rubles. And investments in the development and conduct of tests can cost up to 300 million rubles.^[1]

Notes