Kaliningrad region presented a neural network to fight vandals

As part of the Artificial Intelligence Day, Kaliningrad residents presented the best regional practices of implementing artificial intelligence and the latest domestic developments in this field. Visitors to the Kaliningrad stand were able to get acquainted with the work of the QWATCHER software system - a neural network for detecting acts of vandalism, which was created by scientists of the Kant BFU together with the staff of the regional SBI "Safe City".

A specially trained neural network processes data from Safe City's video surveillance cameras in real time, detects the appearance of new graffiti and other drawings, and transmits a signal to the operator's workplace to make a decision on how to respond to the situation. If the recognition is correct, the operator transmits the information to the Unified Duty Dispatch Service. In 2023, the neural network recognized and transmitted information about more than 80 graffiti that were painted on bus stops, historical and other objects. Since the beginning of this year, the program complex has recorded more than ten such incidents.

Deputy Chairman of the Government - Minister of Digital Technologies and Communications of the Kaliningrad region Levan Daraselia noted the importance of cooperation between the authorities and higher school in the development and application of artificial intelligence technologies.

"Artificial intelligence technologies are actively introduced into our everyday life, they find more and more use in its various spheres. By detecting acts of vandalism at an early stage, the neural network makes it possible to save the budget of public services. This is an example of a successful and socially important application of artificial intelligence technologies. We will continue our joint work with Kaliningrad scientists in this direction, we will use the potential and achievements of our scientific centers," the deputy prime minister said.