INTELLIGENT ANALYSIS OF THE URBAN SOUNDSCAPE

Abstract

For many years, when planning cities, great attention was paid to the visual component: composition, colorist, etc. The sound of the city faded into the background. Sound elements are an important part of the city's cultural landscape. The article analyzes the existing approaches to the formation of the sound landscape of an urban area in Russia and abroad, and identifies the main directions of development. Modern information technologies make it possible to automate the process of assessing the current state of the sound landscape in the city through the use of machine learning, and subsequently manage individual sound elements to form a comfortable urban environment. The authors propose a process for analyzing and processing the original sound of the city using neural networks and studying spectrograms that divide the track into separate components, for example, the sounds of cars, the sounds of a construction site, etc., and also allow you to create a set of sound files that take into account the changes introduced by city planners . The resulting files with urban noise can be used in sociological research to identify the preferences of residents of various microdistricts. An example of the application of the proposed solutions for the formation of a sound landscape is given on the example of a section of the territory of the city of Volgograd.