Students from the Russian University of Transport (MIIT) have presented a unique algorithm that analyzes the noise level in various residential areas. The system creates an interactive map displaying the distribution of sounds across the territory.
Artificial intelligence takes into account the peculiarities of sound wave propagation in the urban environment and complex terrains. It classifies noises into 16 different categories, which helps to determine where the sound is coming from - whether it's a car, a party, or construction work.
The noise map will be available online this fall. The developers are confident that it will become a useful tool for various organizations responsible for monitoring compliance with sanitary standards. These organizations include Rospotrebnadzor, the Department of Nature Management and Environmental Protection of Moscow, as well as management companies and public organizations. In addition, ordinary citizens will be able to use the map to solve problems related to noise in their area.
The project author, student Andrei Maksimov, noted that existing noise maps have their drawbacks. They are either based on rare measurements, which makes them outdated, or on theoretical assumptions, which do not always reflect the real situation. The new algorithm will allow users to independently upload noise data and use certified cameras to measure the sound level, which will significantly improve the accuracy of the information.
Read more on the topic:
AI helps engineers from Russia develop a reusable rocket
AI learns to recognize student performance based on neural networks
AI-powered platform for training pathologists is being developed in Russia