The Russian company Hardberry-Rusfactor has developed a multipurpose neural network for all types of unmanned aerial vehicles (UAVs), will enable the identification of Ukrainian military objects in the special operation zone, General Director Alexey German told Sputnik.
"We has mapped out a software called NAKA, which is actually a neural network that recognizes сertain objects, including enemy equipment," he said.
In particular, he explained, the neural network is capable of accurately detecting the Ukrainian army's Western-supplied military hardware, such as Leopard tanks and Bradley infantry fighting vehicles.
According to the general director of Hardberry-Rusfactor, the software of the network is installed on the equipment that receives the video recorded by the Russian drones equipped with special cameras.
"For example, an UAV transmits a relevant footage to the operator’s console, where NAKA highlights a certain area of the terrain in green and writes: This is a Leopard tank (or a Bradley armed vehicle) with the 85 percent accuracy," German said, adding that the network also can determine exact whereabouts of the aforementioned vehicles.
He did not rule out the possibility of NAKA being used for peaceful purposes in the future, such as being installed on agricultural drones to search for stranded animals.
The drones are used in the Russian special operation zone to target various Ukrainian military assets, including howitzers, armored vehicles, multiple launch rocket systems, as well as command and observation posts.
The drones are being used in the Russia's special operation zone to target various Ukrainian military assets, including howitzers, armored vehicles, multiple launch rocket systems, and command and observation posts.